Brainstorm AI San Francisco 2025

With production ramping up and a new wave of models in the pipeline, Rivian is entering its most critical chapter yet. CEO RJ Scaringe talks about how AI, autonomy, and bold engineering are powering the next evolution of the company and the EV industry itself.

Rivian's Next Charge

Conferences

Intelligent automation promises to transform the foundations of finance thanks to algorithms that can easily see patterns humans can't. Tasks that once took weeks now resolve in minutes. Trades are completed faster than a human is capable. As drudgery gives way to higher-value decisions, how will the landscape change for flesh and blood finance professionals?

Finance in Flux

A generation of Gen Z founders is redefining how intelligence shows up in the world. They’re advancing the science behind how machines perceive reality while reimagining how humans connect, create, and share meaning. By blending research, storytelling, and design, these emerging leaders are shaping how artificial intelligence moves through everyday life. As AI pierces the cultural zeitgeist, they share how they’re helping to shape the future of technology, creativity, and human connection.

The Future Makers

Welcome to Fortune, Tiara. Thank you so much for having me. I am so excited for this conversation. When we were putting this agenda together, someone on the team suggested that we should have the Gen Z perspective, and I was like, yes, I want to be the one to moderate this conversation. So, um, I am a big believer in including youth voices in, um, shaping the future of AI, and you are right there making it happen. So I I'm guessing a lot of the audience are not familiar with who you are, so I would love to give you an opportunity to share a bit of your story. But you are so impressive, you actually, when I was prepping for this, you all, I was like, oh my God, I feel like a loser. You've been named to the Time's most influential teens, Forbes 30 under 30. You're a Peter Thiel fellow, a YC Combinator, uh YC alum, Y Combinator alum, and you're the founder, co-founder and CTO of Chima. So just give us a little bit about your story and how that all culminated into you starting Chima. So I did computer science at Stanford and before that I actually have always been very passionate about research and algorithmic development. So pre um college, I won the Google Grand Prize for seminal research in AI and applying that to climate change. You, you were like 16. I was 1617, OK. And I, I've always been really fascinated about the advancements that we can make with AI models, but also applying them to real world use cases, very much with what we're seeing right now with large language models and generative AI being applied to real world use cases. So I've always been really passionate about this application layer from sort of the model layer to real world use cases and Stanford, a lot of my time was spent at the human-centered AI labs where we actually saw this rich amount of data being applied in different use cases and that really was some of the founding stories of how we started and we operate as an applied AI lab and we are backed by YC, General Catalyst, incredible angel investors, and really Focus on what we call Sam Altman as an investor. So, so what we really do with an incredible number of supporters is, um, focus on how we take these incredible models from different, uh, sort of AI labs but really unlock it in application for different enterprises. Yeah, I would love for us to just spend a few moments, um, double clicking on that. What, who's your target customer and Bring this all to life. Give us an example of what you would do for one of your customers. And so it's really interesting because what we call our target customer can range from medium to a large scale enterprise because what we really see is that as AI models right now from let's say Opus 4.5 to a GPT 5.1 get really smarter, the amount of use cases they unlock exponential. increase, but even if the models stopped increasing in efficiency and improving in the benchmarks right now, the amount of economic unlock and efficiency unlock that they provide is still incredible. And so there's still so many ways that they can be applied that they haven't been applied that even if they infiltrate every part of the enterprise is really incredible in terms of what we really do. I really understand from a very clear use case how a certain function or operation occurs in the enterprise. Think about it as a full back office process and we take it all the way from actually automating that full back office process to end to end redoing it with AI from the ground up. Interesting. So you would come into a potential enterprise and you would like inspect. Workflow and then reimagine that with AI and and it really ranges from being able to apply not just basic large language models through coding agents to also computer use agents. So everything from how do you take the best and most capable computer use agents that can actually automate RPA typical work to actually being able to go fully multimodal. So if you think about this as a Full go to market function being able to understand from let's say the research perspective having deep research agents unlock and fulfill that full end to end function to actually going fully fully multimodal. So what does that mean for uh an enterprises' full campaign as they expand it and enter an entire new market? Yeah, very cool. So you are building in an incredibly fast moving AI landscape and you're doing it super early in your career, um. Without decades of legacy kind of patterns to unlearn, so what advantage does your youth, um, give you as a co-founder? Is it your mindset, your intuition, your hustle, something else? I love that question because I think specifically being part of Gen Z and the sort of younger generation being builders is what we're actually seeing is that most young people that are building these models have an inherent fluency with how they interact with them. We're not. The truth is the younger generation. Isn't adopting AI. We're growing up fluent in AI, so we aren't thinking about coding from scratch. We're thinking about coding with a coding agent side by side, and that fundamentally changes how you write, how you take tests, how you apply to jobs or different applications, because it's not from the ground up. It's actually being able to do that. different models or agents side by side and I think what that really means is that this broad level of use cases and applications we're seeing is really being pioneered by the younger generation. So we didn't have the mobile era as as being sort of builder first, but we do have the AI era which is going to be, if not bigger than than the previous one. I'm gonna go to questions next, um. But I want to talk about the future of work. So yesterday we had a really robust conversation around is AI killing entry level jobs, and it sparked a lot of debate, and I actually want to make it personal for a second. So I have two Gen Z kids. My daughter is 22 and she just graduated from college. He's a food anthropologist. It's a thing. And my son is almost 17. He's a junior in high school. He's my AI guy. He's AI forward and, but he's trying to figure out what to, you know, what to study. Um, so what advice would you have for both of them? Well, I think your daughter is incredible and probably it's not going to take her job because I don't know, but I also think that a lot of the conversation and the anxiety around AI is because often we haven't sort of embraced that it is coming, right? The actual sort of nature. of the future of their work is going to change and what that really means is things like customer support or typical back office jobs are going to change whether they change now or whether they change in a couple of months or years. What that means with the new types of jobs that they take on is something that I think The best positioned to think about very deeply now um but with that being said, I also think that them being able to build in the space is incredible, right? I think they're well positioned to not just join other companies but also contribute to the space because. Never before has there been so many unexplored ideas and use cases with AI that they obviously we've all heard about the one person billion dollar company, um, but what that really means is they can be that, right? And so being a builder or being somebody that's excited to contribute is incredibly exciting right now with AI. Yeah, amazing. Um, do we have any questions? I can't see. Is there a question here? Yes. Charlotte. Hello, my name is Charlotte. I am with Fortune. Um, Kara, I have a question for you that I want your perspective on. We hear a lot in the media world from a lot of CEOs. Like I feel like one of the main talking points is AI is not gonna take your job, but the person who knows how to use AI does. Um, my question for you from your POV is like, where, where does that statement sort of stop and end? Like what is the meaning from your perspective behind that statement and how do you see that like in the future? I love that question because I think. Coming from some of my earliest research has been in climate change, and I think a lot of people are very familiar with the term climate anxiety, which is this idea that there's this movement of climate change coming up and we don't really know what to do, but we know it's coming and nobody is moving as fast to solve the problem. There's a similar thing with AI which is AI anxiety, which is there's literally just In the past couple of weeks been two model releases that have engulfed the benchmarks in such an enormous way that pretty much everything you've ever used AI for can now just be 10x right before Gemini 3 we didn't know what was possible. Now there's a ton of other use cases. So there is a sense of AI anxiety because if you're not consistently ahead of it or on top of it, you're kind of left behind. And the truth is that If that is the case, how can we all stay informed or use these tools every single day so that we do feel as up to date as possible. Now what that means is I'm not saying everybody has to sort of keep up to date with every model release, etc. but that does mean sort of arming yourself with a couple of different tools, whether that be I I'm, I'm not gonna. Sort of name specific ones, but if it's sort of Chat GBT or Gemini, because ultimately there are sort of some main model players are coming out and being able to use them as co-pilots and tools in your everyday life, right, because when Gemini 4 comes out, it's still going to be sort of in the Google interface, right, when the next open AI model comes out, it's still going to be in chat GBT being able. To understand and use these different tools as they expand are really key to staying ahead, right, because being able to use them in different interfaces allows us to become really familiar with them, become really comfortable with them, but also use them as they continuously push different interfaces out, um, but also not having that sense of anxiety. About it because you're sort of comfortable with them. Kiara, um, one of my investment thesis is human centric AI, so AI that creates massive economic opportunity but also unlocks human potential. You did your masters at the human centered AI lab at Stanford. What are some of the principles or values that you, um, kind of explored at Stanford that you've brought with you to the company? I think what's really interesting about this concept of human-centered AI is that pretty much every Interaction with a model, whether that be an image model like nano banana or um a sort of large language model is really rooted in human centered AI. And so we see incredible benchmarks, we see um sort of the next model improving on the on the SWI benchmark and coding agents, but really if you take a step back, what makes specific models feel different than the other ones you see this. Very clearly in image models, right? People see and and work with nano banana and multi-turn edits, and it really is because they see something in the image quality that's very different to a seed dream, right? Um, it's really diff difficult to measure the quality differences between these models and, and why that's the case is that as models and use cases and efficiencies change, the the the key differentiator. is taste. Gary Tan from which is not captured in the benchmarks around accuracy, right, right, and we see this with design. If you've ever sort of coded with a coding agent, you know it loves a couple things. Sparkle emojis is one. It puts it all over front end UI. It loves sort of shades the end sort of components in a way that um you, you know something is vibe coded if you've ever sort of worked with a coding agent and the. The the key sort of differentiator is what makes it taste and feel different, and that is human centered, which is what do people actually want to see in an AI application or an AI use case. Yeah, I would love to see more benchmarks that kind of capture that aspect of these models like the design principles and kind of the the human aspects of it and I'm involved with this MIT group. It's called Advancing Human. With AI and they are actually trying to collaborate with some of the big kind of the open AIs of the world to build these benchmarks. So it'll be exciting to see, yeah, I, I think as models code more on the front end and, and less on the back end, that's, that's going to be so important. Do we have more questions in the audience? OK. What is, um, the biggest misconception about the Gen Z, um, usage of AI? Like how do you guys use AI? I think that or think about AI. I think that the biggest misconception is that young people are using AI to not think things through. I think that like as a shortcut, yeah, I, I think that really intelligent Gen Z, um, individuals are using it to think even deeper because you can run. A deep research report on let's say parallel AI that goes on for 34 hours and what are you doing when you actually have that report on your fingertips? What does that unlock for you if you're looking at the financial market, if you've done a deep dive on your favorite sort of article or topic, what that unlocks for you and I. I think people think that instead of you doing all that work, you've now sort of handed off the cognitive load to an AI model, but really what does that allow you to think differently about and think even deeper on and I think that's the biggest misconception that people have about young people and their use of AI. I, I know, so impressive, right? So cool. So cool, um. I have to ask you this. You're a Peter Thiel fellow. Did you consider dropping out of school to start your company? Please say no because my son may listen to this conversation and. And, and I do want him to, no judging, but yeah, but I, how did you think about this? I, I think what I would say to, to, to that point is that the community that, that brings people together in places like Stanford is incredible, right? Just because, um, as we're seeing sort of model or AI applications proliferate, where it really comes back to our previous point is how people come together and build really unique things. And, and I think that was key to what is a sort of Stanford college experience. And definitely through my experience I think for his use case or his application, I think that's really dependent on on sort of what how he sort of sees himself contributing to the space building or even working with different AI models. I, I think this is such an important point because in this age of AI, I think human connections are going to be even more important and, and, and experiences and spaces that allow you to build these authentic human connections are going to be so important. OK, so you have a lot of older people in the audience today. Um, what's your advice to all of us? I would love to say that the sort of you can be honest I'm OK when we sort of have a conversation about younger people using AI models and and being really comfortable with them, it's just as important as the older generation using it because AI fluency is just as important for people that are already in the workforce. not going to teach it to younger people to use AI. And the truth is, if, if there's this sort of anxiety about it, we are just at the beginning. The models right now are as dumb as they are ever going to be. What that really means is that when you look at a couple of months down the line, we are going to be in a very Different landscape and so being able to be really comfortable with that and having your core tools that you use um and and and get comfortable with that is really important because we're sort of at the beginning. It is only going to get faster, more advanced and more intelligent each and every model from from here on out. So being. Really comfortable with that is, is no AI anxiety, you guys. Kira, thank you so much for this wonderful conversation. It was great. Thank you.

Telstra and Accenture are reimagining what enterprise transformation looks like in the AI era by joining capabilities, sharing accountability, and accelerating innovation responsibly. Their joint approach offers a behind-the-scenes look at how leadership, a robust Data and AI strategy, and an openness to change are driving Telstra’s AI-first ambitions at scale. What lessons can other enterprises learn as AI becomes foundational to how business gets done?

Reinventing the Enterprise

Investors are pouring billions into robotics amid the AI boom, but only a handful of startups will win in the famously capital-intensive and technically complex robotics space. Sequoia’s Stephanie Zhan and Skild AI’s Deepak Pathak unpack the race to build a general-purpose “robot brain,” separating what’s hype from what’s quietly transforming industry. Then, watch the future in motion as Skild AI’s robots learn, adapt, and perform live on stage.

Brains for Bots

Stephanie, Deepak, thank you for being here. We're so excited to be here. Thank you for hosting us, Sally. I think we should start with the biggest central question right now. The AI boom has facilitated a lot of hype around robotics, but is robotics overhyped? That's a very good question and a very natural one. Like if you look at the whole history of AI right from 1954, from the days of Turing, robotics was the first application that AI was basically founded for, you can say. And if you look at all over the years, We all have seen impressive robotics videos for decades. Now, you may not have seen, but they are super impressive robotics videos even from the 1980s, 1970s, but still no robots around us, and it always feels like robotics is coming in the next 5 years, next 5 years, next 5 years. It has been 70 years. So is it overhyped? Is it not overhyped? It's a very genuine question. And one thing to look at here is the way robotics is done today, in the last, uh, I would say even 3 or 4 years. is very different the way it was done for the past 65 years before that. And that's a big change. Now, what is that big change? In one line, the change is Things in robotics used to be driven more by human intelligence. Like somebody super smart would look at this task X Y Z and have the robot preprogram the robot mathematically to do it. What has now changed is that these models or these robots can now learn from data. So that has shift from programming something to learning from experience. And that is what this new wave is about, which is so different about robotics. And this is, we know this approach generalizes from all the results we see around LLMs, VLMs, language models, and all that. And that is one big change from the all 5 year chunks before that and this 5 year chunk what we are seeing now, Stephanie. So to put that in perspective, Deepak has this wonderful saying that he likes to to repeat, which is that what looks hard is easy, but what looks easy is really hard. And I'm sure that all of you have seen the videos on X that go viral. It's a robot doing a backflip, a robot running in a straight line, a robot folding laundry or doing some kind of household chore, and all of that looks impressive in the abstract. You know, in a vacuum on X, it looks amazing. But it actually requires expert discernment. And I think that um this is what Deepak is getting at. What you don't actually see is whether this robot was trained for general intelligence. Whether it was able to do that task, actually interacting with that environment, or actually capable of executing that task if something in the environment were to change. A small tweak, for example. What Deepak and Skilled are capable of doing is that they're building this general intelligence that enables robots to do anything, anywhere on any robot hardware. And that's massively transformative. It's similar to what OpenAI and Anthropic have done. Um, in the world of digital intelligence, but instead they're doing it for the world of physical intelligence. And so, you know, the challenge I think for you as a consumer of all these videos is to really discern what is real and what's not. And even amongst the videos that go viral online, there is massive overhype, um, but there is. Real progress and real trajectory that we're working towards and that's what I'm so excited about with skilled. Well, this is the point both you and Deepak have been very, very clear with me about, which is that it's actually a lot easier to program a robot to do a backflip than it is to get them to climb stairs. You're smiling. Why? Please, I think we should maybe show that video. Let's do it. That, can we cue the parkour video? Yeah, no, that's, that's not yeah, here we go. The issue is I, I sent Ellie multiple videos to choose from with no clue of which one to choose, and she chose this one. Now this, now it will look super impressive. Wow, the robot is doing this. I cannot do this. Uh, and it, it is like. And it can keep on going. It's funny, right? But one thing you would notice here is There is very little interaction of the robot with the world. Apart from ground, the robot is not interacting with anything other than ground. Maybe there was one block in there to jump over that block by putting their hands over. Now if you think about this, so this going deeper into this, uh, OK, I'll let the video play. Nobody will listen to the voice. Well, I was gonna say the other thing about this too is, in part, a huge part of what you're doing is around a robot brain, right? What does that actually mean? Yeah, so one thing, uh, like, uh, if you think about this, this component, hard, easy, easy is hard, right? This was, this is not my quote. It's from Hans Moravec. He was also a CMU professor, uh. This was the culmination of 30 years of work from the founders of AI from 1950s to 1989, and this was their whole conclusion, heart disease is. Now what does that mean? It means that, for instance, uh, If somebody can do a backflip in humans or if somebody can throw a ball into a hoop from very far away, you would pay millions of dollars to watch those people on stage do that, right? But then if somebody goes and opens a door or climbs a stair, you would not be surprised. Why? Because every human can do that. Now if you think about both these tasks, what is different between the two? The difference is interaction with the world. Backflip is full control of your own body in a free space, right? Which is exactly what computers are good at. If you tell a computers everything about you, about everything controlled, they're good at doing that. One example is airplanes. We have had airplanes for almost 100 years at this point, right? They fly in the free space, so there's no interaction going on. But when you talk about picking up a bottle, a glass of water or going on stairs, what that involves is using your vision. To immediately to continuously correct where you keep your next feet on the stairs. Any one point here and there and you and you fall backward all the way. And that interaction is what is the root reason for human gender intelligence. It's called sensory motor common sense, which you don't appreciate because every human, almost every human has it. Yeah, I was gonna say when you put it that way, walking upstairs is an incredibly high. endeavor, um, I mean, Seth, when you're thinking about this, you know, Sequoia is famous for being interested in big markets. What do you see in terms of use cases, market opportunity? Well, I would highlight a couple of things, um, you know, if skill succeeds, we completely transform what robots are actually capable of today. The market's massive already, you know, you know of companies like ABB. Kanu, Kuka, Mitsubishi, Kawasaki, that alone, just on the industrial side, plus all the consumer folks, that's massive. But they're still constrained by having robots that are only built for specific things. And the real potential of what skill transforms is opening up the possibilities that are otherwise not possible for us to achieve. So if you think about Commercial settings, security and surveillance, robots always fail at stairs, elevators. But if we're capable of actually producing robots that can do all those things, we can tackle security and surveillance. Take industrial, all the different types of tasks that that would otherwise fall under edge cases that robots still fail at. If skill succeeds, they can tackle a whole host of industrial use cases that often pose dangers to humans because of the materials and cramped spaces that you now have to deal with. And then finally, in consumer settings, Um, you know, you see all these household robots, but they're only capable of doing one thing. It's your one vacuum cleaner, your one dishwasher, whatever it is. But if you're if we succeed at building the general intelligent robots, you finally have consumer robots that can actually tackle the whole host of household tasks that you now have. And so the opportunity here is, it's really all of the unlocked possibilities. The a parallel here is, you know, if you had thought about OpenAI and Anthropic back in 2016, it would have been. Hard to actually understand what the scope of market potential was because we were still living in this vertical AI world in software, but the real market for an open air anthropic is the market potential for all of digital knowledge work. And in the same way, that is what skills unlocks. It's the market of all of physical labor work, all of blue collar work today. And not only that, I mean that alone is transformative, but what's even more interesting is we can get all of this at accelerated pace and at a fraction of the cost. Because you have something that is generally intelligent, you no longer have to program these robots for specific things that take a long time to program. And they're only available at 6 or 7 figures today. But if they succeed, you now have this generally intelligent software that can run on any robot hardware for an order of magnitude smaller in cost. And so that's what's most transformative and excites us about skills. We're going to turn to the audience for questions soon, but one thing I wonder about is, um, especially with the influx of capital moving into this space, is it a winner take all market, a winner take most market? I'm curious for both of you. I think this is a very, very good question. Uh, it's always hard to predict about the future, but in, in my opinion, I would say that the answer is very technical. Why is it technical? Because in robotics, if you see, what's the main issue? The main issue is data. Like if you see the language model land, whenever these language models, whenever show some result, any company, let's say open, they release some result. You see 5 of the labs almost reproducing the same result in a matter of months or a few months. Why is that? Because data exists for language models on the internet, for vision models on the internet. Now, whether you can use it or not, it's a copyright issue, it's a separate issue altogether, but it exists. But for robots, there is no data on the internet. So how do you get, how do you begin? Then uh training all these models to start with. Like the huge success of AI is driven by big models, big data equal to success. And we are missing big data in robotics, which is where the key part is. Right now we have to use alternate sources to somehow bootstrap robotics to get it out there in the field. But once it goes out there in Whichever company gets out there first across huge diversity of tasks, that creates sort of a data flywheel from robotics, right? And the interesting part is that's different from self-driving though. Because in self-driving, people still drive cars because cars help them go from point A to point B. You don't drive a car to help self-driving. That's not your goal. But in robotics, so robots have to be useful for something to begin with. But once, if somebody deploys these robots in large numbers across large numbers of tasks, Then you create the data flywheel, and that becomes your mode in some way to go forward. And when that becomes your mode, it almost becomes more like rich gets richer. Like, the more data you have, the better your models are, and the more data you get. So it becomes harder to catch on. And this is a classic trajectory for markets. Which are not heavily democratized like it's monopoly duopoly, triopoly, I don't know. So the winners will really be the winners. Yeah, exactly. What's your take, Steph? I agree with that. I think that the market traditionally has been very segmented based on different industries that you specialize in, but the reason is that you specialize in making robots for automotive, for industrial, for something else. If skills is successful, then they completely create this new category above all the hardware players. They become the generally intelligent software that powers everything else, and so it has the potential to be more winner take all. And I would say that I deeply believe Skilled is the leading frontier lab for building all of general physical world intelligence. They're leading in pushing the frontier. This is just a snippet of what you've seen. Uh, there's much, much more that you should check out on there. There are some, there is a video that involves a chainsaw and the robot brain the legs off a robot that where the robot actually exhibits emergent intelligence and generalized intelligence, so you should go check that out. They're the leading, they're leading in pushing the frontier research. They're also leading in terms of actually getting the. Robots out with real customers in in the real world and then I was the third thing I would say is they're leading in terms of becoming the talent magnet for all of the most elite talent in this field. And so it's been really exciting just 2.5 years since our founding, and they're on an incredible trajectory and, and much, much more to come. I think that it's moving at an accelerated pace, um, and it's, it's exciting to see. I would say you're a little bullish. I would add 11 thing to the argument about uh this monopoly versus like uh uh uh multiple companies. What I meant this argument was more for the software, right? But if this is of the hardware, it's very rare that people would want to, everybody want to buy the same hardware. So hardware is very, very difficult, unless you have, unless like Nvidia, like some chip or something. But if you're, if you the end hardware, People just want preferences. If you, if your neighbor has some hardware, and your other three neighbors have the same hardware, you would want something more unique for yourself. And so hardware, in the hardware side, you would see multiple companies, because that's how it is. Multiple form factors, different designs, different uh uh appearances. What, what I made the argument about, uh, this data, that's more for the brain part. Kind of like phones with Android kind of situation. Questions. We have time for one question. Oh my gosh, there are so many questions. How do I pick? Whose hand was up first? OK, I'm trusting you. I'm trusting you over there. Um, Deepik, one of the things you mentioned was that every year in robotics, it feels like by next year or the following year, we'll have something that's actually more mainstream. What do you think is gonna be for consumer applications or for people in this room, the time period over which we'll actually start to engage with robots on a daily or weekly basis? Yep. Instead of giving time frame uh for consumer like that you can buy or not, I would say one thing before you see robots around you in houses, which almost seems like the, the main narrative these days. You must be, you must have to see robots around other areas like when you go to grocery stores, when you go to hospitals, when you go to uh other more structured hotels. We are not seeing that right now. I think first that wave has to come, and then once you are already familiar with robots around the world, you're seeing it in real life, then you will say, OK, I have to buy those things now for my personal use case, the same way how computers began. First, computers were used for professionals. They were more for companies. The idea of personal PC came much later once people got familiar with them around it. The same timeline you would see in the shorter term. What we will see next is More and more robots are deployed in the enterprise scenarios, which for the jobs, which are not where not everything is known or not pre-structured, where humans are involved. Let's say a human is doing some task on an assembly line, robot works alongside that. And that is not possible today. Today, robots that have to be only in cages and things like that. So you start from those areas, which is where, where skill is beginning as well. We are already deploying in a variety of areas, variety of, uh, Um, factory setups, but all those setups are the ones where there's randomness involved. And they act as bootstrapping, uh, uh, platform for us to go to the next stage, which is more unstructured, which is more like hotels, hospitals, things like that. And then they eventually stage is consumer. It's the same task everywhere, same skills everywhere. So it's, uh, the ordering has to be some somewhat like this. And I'm going to take it home. I'm sorry to all of you who had questions though, um. There is a lot of anxiety around robotics. As much as there are all the fun videos, there are also concerns about killer robots. There are concerns about robots taking, taking jobs. What can our robotics future look like? I'll start with you, and then we'll end with you. Well, I would say that you know there's a large labor shortage in a bunch of different areas. Um, there is incredible pain that people feel in just the churn of labor in different industries. So that's an area, both of those are areas where I think a lot of the services that robots will now be capable of offering will first start to tackle. It's the places where there are already pain, um, a high urgency, um, and desire to adopt fast. I would say there are 3 essays of robotics future. Three Ss, OK. The first S is safety. Uh, now, this is not robot safety. If you look at the human jobs today, right? We are right now sitting in this room, we are a bit privileged to be working here. That's our job. But if you look at, there are still so many jobs out there, where every day people go to work and they risk their life. Either for a short term basis, some risky jobs in a, in some scenario, or they receive their life from chronic issues from like continuously bending, you know, in a position and doing work. But people have to do this work to make the society function. So those are the jobs where you definitely need uh robots. It will only improve the quality of life of the workers who are working, doing these jobs. It's the first, safety of people are doing dangerous jobs. Then 3rd, 2nd, 2nd reason is If you think about jobs. While this is true, the narrative is that robots will take jobs, but if you look at the employment statistics right now, there are almost a million more job openings than people are available to fill them. At the same time, there are so many people who don't have jobs. What is the discrepancy? It's a discrepancy of It means that people want to do certain kind of jobs. So there's a huge, almost a million gap, and it's only increasing, uh, after COVID. It has really widened. So people want to do certain kinds of jobs but not other kinds of jobs, but they have to be done. Like today it's a huge labor shortage for blue collar, uh, workers for big factories, big, uh, big, big corporations, and it's totally the opposite narrative. a million jobs short, that's the highest, uh, we have seen, sorry, million labor short of the jobs available, right? Now the third aspect is, let's say we all, everything works well and robots do really well and they get deployed. Then the 3rd aspect is more of a task for all of us. It's a social aspect. It is true that when robots become really good, uh, in the future, there has to be some work around how do we handle the job aspect, the salary aspect, and what people do. And I believe there is a local minimum. It is a social issue, like, so that not only few people are benefiting from the same this evolution, everybody is benefiting because there's abundance of everything. But if we go beyond that, then there lies a scenario, a good scenario where everybody's doing things that they like, uh, work is more optional, and they are doing things that they enjoy instead of being forced to. And in that scenario, there are so many robots walking upstairs. Deepak Stephanie, thank you so much. Thank you, Ali.

They’re tiny, quiet, and smarter than you think. MJ Burk Chun reveals how Serve Robotics’ autonomous fleet is mastering urban navigation and reshaping delivery through sensor fusion, real-time learning, and sleek (plus adorable) design. The session concludes with a live demo of the company’s latest robot in action.

Bots on the Block: Your Lunch is Learning

Alright, good afternoon, everyone. Thank you so much for sharing your afternoon with me. Um, I'm gonna talk a little bit about Serve Robotics and geranium is gonna help me here. Um, so just starting off, uh, a, a couple of slides. Um, I don't know if it's come. Oh, there it is. I totally missed it. So we are an autonomous, uh, delivery robot. Start off within the Postmates, uh, tech innovation lab with the question of why move two pound burritos in 2 ton cars. Really, there was inherently in this, uh, Postmates data, and we were, you know, born in, in SoMa in San Francisco, was this a concept of why it was so much cheaper to get something from Chinatown, or sorry, from China than from Chinatown, right? Like this is just like a known thing. Um, and whether or not you actually can find yourself a 2 pound burrito or they're a little bit lighter, it's still an inefficient means of transportation from a delivery perspective. And we looked at all of the delivery data, and it's a three-sided marketplace. You're getting drivers to agree to a job at a restaurant to a customer. And part of that driver pool, they're really interested in maximizing their time. And so what they weren't interested in were short distance deliveries with low tips, especially in dense urban areas where now. You have to find parking, maybe you're double parking, maybe you're risking that fine, um, and you're jumping in your food may or may not be ready, and that's eating into your earning potential for the, the much more lucrative job. So that's how we started dreaming and evolving robots. So we actually, uh, partnered with San Francisco, um, had a pilot program. And I think week two, we had that first little iteration, um, we had someone's phone. On RC car and drove it up and down Market Street and said, alright, what are we gonna learn about what it means to be on the sidewalk? Um, then we had another iteration where someone was like, oh, to Hurt Locker. And so, you know, we, we were like, oh, OK, this is where design really started to matter. We're talking about technology and embodied AI in our world. Um, that's, that's not gonna win anyone favors, um, and so we had what we affectionately called truck, and then we evolved to more of this robot, uh, form factor based on a shopping cart because it was familiar and we really had this idea that you wanted to be welcome on the sidewalk. And so, um, really, what are we trying to do? Safe, socially aware, and this is where the people part really matters, robots in a dynamic world, because people are unpredictable. Our cities are very unpredictable and diverse. And the theory, uh, this is where the, the AI portion really comes into play. We have the concept of like identifying people, uh, very traditional method, um, understanding what's the, what's the wall, what's the ground. Uh, what's the pedestrian light signal? What's the status? Can the robot cross the street? Not a joke, a real thing. We're like, oh, let's let's spend lots of time talking about how do we know that a robot is safe enough to cross the street. Um, robots have nightmares about cars. Cars are also very scary for robots. The intersection is actually one of the most dynamic places in our cities. You've got high velocity moving vehicles. And you've got a robot that's trying to share the public space as it crosses the street. Um, so it's a very, uh, you know, autonomy and hardware bound problems. Truly like where AI in 2017 was all about. And then we, we deployed in Los Angeles. We're doing deliveries for the very first time with robots, and the reality was actually even more dynamic than we expected. Um, that's one of my favorite photos. Uh, the robot was thoroughly confused about the pet baby goat. It's just like unidentified does not you know, like see the person, but like what is this, um, that was, that was one of the moments where like, oh this is this is gonna be an interesting long tail problem um sidewalks very different from simulation. The discontinuous tree roots, uh, the infrastructure is not quite there, uh, and, and so we started learning a lot more and evolving the hardware to be more reliable, um, and, and really to really traverse the neighborhoods and so one thing that's really interesting about this like sidewalk experience for us. And looking at this room and some of the conversations that's been happening really to me is all about how the future really is already here and I think many of you are involved in that particular version of the future. It's just not evenly distributed. William Gibson, uh, he has a great series of of books about the near future in terms of, of speculative fiction. Um, we're adding new cities all the time and so we're adding our, our. You know, flavor to these cities, um, in terms of what it means to, to live in a city. What's really, really critical about being physical AI in these cities is we have to be welcome. Robots don't have rights to be on sidewalks. People do. And so most of our design decisions, engineering decisions, and safety decisions really are around how to be conservative, but effective in a dynamic environment. And it is also quite amazing how each city expresses itself in the way people walk, might be a little different. Not just the sidewalk infrastructure, but also how people drive. And so one of the ways that it's a delivery business, it's all about being on time and reliable, one of our things is we've invested in routing and we learned that the conservative routing that we've optimized for in Los Angeles in, uh, very high traffic, high speed intersections doesn't quite hold up in Miami Beach. Folks are gonna cruise down Miami Beach. They're not racing to make that right hook and from a routing perspective. Might not make sense. And so we're learning things as we scale and really, once again, it's back to that human element. What makes cities dynamic, vital, interesting, is the people, what's really specific geospatially, it's locale, it's community. Um, and so we're learning a lot about how we can build better experiences for people, leveraging that learning, and it wouldn't be possible without some of the advances in sensors, sensor technology, um, and autonomy, specifically around end to end AI, uh, we're able to take. LiDAR data, camera data, movement, map it to GPS coordinates in the cloud. We've actually taken, um, we're taking all our millions of miles. We're on, on our way to have one of the largest AV fleets in the world. Um, and these are images, uh, and really maybe not the most exciting images for some of us. Um, we care very much about the sidewalk, the sidewalk quality, whether or not an intersection is missing curb cutouts, uh, so it's, it's a lot of data that we're collecting about the status of the city, the state of the city, and then we're leveraging that power not only to train. The um algorithm so that the robots are behaving better but we're actually sharing the data with cities so that they can address things like that little yellow circle it's a little hidden pothole and we're changing the city infrastructure by improving it with the data that we're collecting by doing deliveries on the sidewalks. Um, and I think the last piece, go back to it, is that embodied AI perspective of always delivering delight. Um, I think this is where I wanna maybe be humble and say more people will walk next to the robot, drive a car through an intersection, than we ever get a delivery from a robot. So if you're, if the robot's not delighting someone. Not providing some value to the community, then it's a missed opportunity to really build that future that I would hope we get to all live in. So, with that, uh, thank you for allowing me to share our story. Um, geranium is also very excited to participate in. So, um, with that, I think it's it's Oh, Exit to your left. Thank you again.

Almost a year into his mayorship, Daniel Lurie is navigating the challenges and opportunities of keeping a leading tech hub on top. In this discussion, he'll talk about the transformative role of artificial intelligence in city governance, what role the tech community plays in the city's future, and how he plans to bring more business to the city by the bay.

Leading San Francisco: A Conversation with Mayor Daniel Lurie

Thank you so much. Hey Daniel, how are you? I'm good. How are you? Good. So you're a year in, almost, not quite. January 11 months. Yes, sir. Are you tired? No, no, not at all. No, I don't. That, that doesn't enter into the vocabulary. Impressive. I think he's part robot, folks. OK, um. It's been an extraordinary year. Um, you have, I, I, I believe at this moment, last I checked, 7 71% approval rating. That's pretty good for anybody. That's pretty good for anybody who's a year into anything, I would say, um. And you avoided a Trump immigration standoff in this city. Um, you have dealt with a litany of civic issues. What are the biggest challenges you face in your opinion as the mayor of San Francisco? I think when we started, I think it was, um, making sure that San Franciscans knew that this city, uh, when we are at our best, is still the greatest city in the world and turning around sentiment, turning around that hope and that optimism, that was our big challenge. We had so many voices from outside of this city. Saying and defining who we were, that we forgot that we are the ones that are going to define San Francisco, and I think we've done a really good job of that. I think those that stayed and fought for our city, that didn't flee, that stayed and fought and invested deeply into our city have seen huge payoffs. I mean. Everything is rushing back in. I mean we've got retail rushing back in. We have people buying office space and buying hotels, and I think sentiment was the big challenge, and I think we've overcome that. At this time last year, 25% of San Franciscans felt like we were heading in the right direction. Today 62% of San Franciscans think we're heading in the right direction, and that's where we can now build from and now we have to execute. We've got to execute on public safety, which we are. Crime is down 30%. We've got to execute on the behavioral health crisis and getting people off the street and into the right kind of treatment beds and recovery beds. And then third priority, we have to strip away the red tape to help our small businesses, our restaurants. Our our galleries get up and running more easily and have a City Hall used to be kind of an opponent for our small businesses. We want to be a partner and we have changed that drastically as well, and we can go into that. But I think the biggest challenge was sentiment. We've gotten there and now we need to go execute, keep people safe, keep our streets clean, and, and create the conditions for success. Well, the sentiment thing is very interesting because you started off talking about everyone else's sentiment outside the city. But in fact it seems like the biggest nut to crack was San Franciscans' opinion of themselves. Absolutely you have to start. You, you got, you gotta love yourself before, um, anyone else is going to love you, uh, and this is now a mental health, yeah, we're gonna, well, we need a couch, um, and I think it's really important that San Francis, you know, we, we lost a little bit of our mojo, you know, we, we also took things for granted. Granted we took our restaurateurs, we took our entrepreneurs, we took our business community for granted we said we can just keep punishing you, we can keep being, you know, uh, a an opponent of yours and you're gonna stay well that didn't happen. People fled. Um, I wish they hadn't, but I understand, uh, to a degree why people felt like this the elected class in San Francisco took people for granted. Uh, from our artists to our restaurateurs to our entrepreneurs, and we're not gonna do that again. And so, uh, this is the greatest city in the world when we're at our best, and I think people are starting to see that again. Yeah, and you were born and raised here, so you really mean it, born, born and raised. So, so this being an artificial intelligence conference, I do have to ask where. Among these priorities that you had and please tell me if they've changed, by the way since you took office a year ago, what, how are you using technology to address them? Are there areas where. It's a lot easier or more difficult. These are big entrenched problems. Probably every major urban mayor has to contend with them. Yeah, I've been clear for now the last couple of years from the time I started my campaign till today, uh, with you here, public safety is my number one priority. Nothing else matters if you can't keep people safe. And so with crime down 30%, uh, citywide crime is down 40%. In Union Square in the financial district where we are right now here in San Francisco and why is that? Is more bodies on the street? I mean, what it's, it's we're still short staffed and the police, but we've, we're using technology, we're using drones as a first responder. We're using, uh, license plate readers called LPRs, uh, so that when something, uh, uh, an a criminal act is committed, we can get a drone up quickly. We can track people. Can do it safely so we don't have to have officers, uh, you know, flying through the streets because that is where accidents can happen and we are catching people now when they commit a crime. So we have, we just launched, uh, relaunched this thing called the Real Time Investigation Center called Arctic here in downtown San Francisco, uh, you know, think the television show 24 where we are. Really able to watch and see what's happening and keep the residents, the business owners, the small business, the retail shops safe, um, and so we're seeing a huge precipitous drop in crime. We are looking at the lowest homicide rate right now since the 1950s in San Francisco. Uh, we are an incredibly safe American city. We still have issues. We still need more officers out walking the beat, so we're going to continue to recruit officers. We have a 40% increase in people wanting to be SFPD now. More police doesn't solve all of our issues because we still have a behavioral health crisis on our streets, so we're trying to get people off the street and into the right kind of beds. But you know the Arctic is one example of technology being used. On all of our fleet of cars we're putting sensors on them to make sure that they're being used efficiently. We know where they are. Our homelessness response, we're managing what beds are available in real time. We're doing a better job. We're rerouting calls, 311 calls, streamlining it using and. With the help of AI and AI agents, and, and so, uh, but we're in the early innings of that, as I know, of course, many people in this room are too. So let's talk about the tech community. You mentioned downtown. That's a big priority for you. You, you, you, you said in previous interviews that, you know, unlocking downtown is unlocking the whole city, right? And so goes downtown, so goes the whole city. So. You know, the tech community has a, shall we say a lot of opinions, yes, uh, about San Francisco and in particular the downtown experience you've been trying to work with them as you said just just earlier a few minutes ago, um, trying to bring them in as open arms, trying to bring them in as partners, um. Opinion sentiment is still divided, right? Some people still talk about SF as it was. Some people still see it as the tech capital, the beacon on the hill. What are you doing to continue to work with them? You're deploying all this technology for your own needs and your citizens' needs, but what are you doing for the tech community to bring them back to downtown, to have them to, to, to make San Francisco what people think it is? Well, first, we are. Working with the community, we work with the business community, we work with the tech community, the tourism community, we want anyone that is home based here, and that is a lot of incredible AI companies that are here, we want them back in the office, and we don't actually have to have this discussion about whether or not it's 3 or 4 days a week in the office when you talk about some of our major companies anthropic, OAI, they're like. 3 or 4 we're in the office 7 days a week and so we're getting people back in the office because we're creating the conditions that people want to come back downtown. Clean and safe streets. I know I sound like a broken record, but my job as the mayor of San Francisco is not to tell people to be in the office 5 days a week. It's to create the conditions so people want to be in the office. So Muni, our public transit, is running efficiently and safely and is clean. Our streets feel safe for people to walk down. You go to Union Square now and it is bustling. A year ago it was not, uh, and so, uh, we, we want their input. We've started something called the Partnership for San Francisco modeled after what New York City started in the 1970s when they were going through a real rough. Time there we launched that there's 35 business leaders, not by the way, not just tech, but tech is engaged and involved, but we have tourism leaders, we have, you know, Visa, we got the San Francisco Giants, we got Ripple, we got, you know, all the companies around the table, and we want to hear their voice. And yes, everybody in San Francisco, it's not just uh tech that has opinions. 840,000 of us all have our own opinions. And I welcome that and that's what makes us such a unique and and great city indeed, um, I wanna come to the audience for a question in a in a minute, but I, I wanna ask you about, so we are both California residents, you and I, um, there have been a flurry of AI laws principally at the state level going on about all that sort of thing. I can't help but imagine that you feel like you're in it whether you are or not as the mayor. Of San Francisco, what's your thought on where the state needs to be in terms of AI policy? Should it be pushing aggressively, you know, we're seeing a fragmentation across the United States in terms of what the states want to do. I mean, listen, I was elected to take care of San Francisco and San Franciscans, and yes, the companies are here. I have not weighed. In on AI policy as of yet, I do think that if you do something that is just focused on San Francisco, we've seen it companies can move to 2 miles away in South San Francisco. They can move to Brisbane. So I'm always a believer in making sure that we have regulations and guardrails in place for this incredible technology that The companies here are unleashing. I think any time you do something that's just city specific or maybe even state specific, you get into a territory where you drive people to other locations, so. Um, not that we shouldn't be leading, uh, not only with the technology but also with, uh, regulations, but I would love to probably see a more, uh, broad based approach, uh, uh, from, you know, our electeds in, in, in the federal government, but, uh, I think we should be always pushing, and I think that's what they're doing in Sacrament pushing pushing. All right, let's go to the audience for a question. Oh look, they're timid after lunch. All right, I've got another one for you, sir. So, oh there's a lot of people we talk a lot about this, yeah, now we, now we can see all of you with the lights up. All right, great, um, you, we talked about the citizens, um, we talked about the tech community. I wanna come back to the citizens. How are you using technology to actually elevate. The non-tech community, the rest of this city, which you, you know, of, of which you're also an elected official representative, um, are there things that can be done? Yes, you talked a little bit about like homelessness and getting folks off the street, but I'm, I'm also thinking about upskilling and. Just elevating the entire city instead of just 11 thing that is around permitting, uh, right now we have a great and good looking permitting office where you can come, you know, to get, you know, your business license and, and, and, and to open a new restaurant and you, but you, you, you, you get in line and you start realizing you gotta fill out, uh, paper forms using. Pen and paper and you go to the fire department at one window and you fill out one form and then you go to planning at another window and then you go to the Department of Public Health at another. I mean those days have to be over um we have we're we're streamlining that we want you to fill out one form on a tablet or on a computer and it goes to all eight departments that oversee, you know, approval process that's right uh and so we've created something called Permit SF. We had a 100 day sprint. We had a 200 day sprint actually later this week we're unveiling some, uh, new. easements around a 300 day sprint just making things easier for the residents, for the small business owners if you want to put new stairs in your backyard or if you want to just redo windows, we have to make things easier for the residents of the city. That's what. Technology is meant to do. It's meant to make our lives easier. We are still living in, in an age where we are not fully utilizing what is right here in our backyard. So we, we are through Permit SF trying to make life easier for the residents of San Francisco. Yeah, I mean, it takes a special person to get excited about. We've got a good looking permit office, you know, um, and, and but we might have those folks here in the room. So, so, uh, as a homeowner, of course I thank you for that. Um, you actually, let's talk about Permit SF just a little bit more. So you, you, you were embroiled in a little bit of a scandal earlier this year around the selection of the technology for that initiative, um, YouGov, I think it was, or OpenGov, excuse me, OpenGov, um, and, and you were criticized because, you know, folks thought that it was not the best tool available, the, the, the selection of that technology was not. Not it and they were, uh, my understanding is your family has ties to that software or something like that, um. It's past you now, but what did you learn from that experience? Uh, well, listen, um, I don't have ties to it, um, but, but let's, let's be clear, uh, when you want to change a system, a legacy system where you have 8 different components to something and you want to streamline and you want to. You know, make life easier for 840,000 people, but change is happening in a really large bureaucracy. You're going to get pushback, and I get it. I understand people are scared of change. I, I am not scared of change. I am not scared to push the envelope and to make sure that we do what's right and make sure that we have more efficiency, more effective government for the people of San Francisco, and there's always going to be pushback when you, when you do that. I think if you come back a year from now and you ask me how has this implementation gone that that'll be the test we're in the midst of of overhauling this there's going to be pushback. I get it that's cool, um, but, uh, it'll, I, I want people to test me on whether or not the the software works and makes people's lives easier, and we're not there yet, but we'll get there, right? I, I believe this is the first time a robot has ever crashed the stage. What's going on at. Brainstorm, um, high germanium. I'm not sure what to say right now. Ah. Could you hold this for me? Yes sir. Thank you. I get to see what questions are coming up next. Very funny, germanium. Thank you. You can keep it. Thank you very much. All right. So let's, let's wrap it in. Come have a seat. Um, the, the, Let's talk about the red tape a little bit. Um. They're so cute, aren't they? Oh yeah, yeah, cool. I didn't know they could parallel park either. Fantastic, um. There you go, um, the, the red, uh, uh, tape of, of, of San Francisco is well known. It is also hardly unique to San Francisco. My hometown of Philadelphia also deals with this problem. My, my residence in, in Long Beach, California, same deal. Like every, a city of, of significant size has bureaucracy. To contend with SFS is particularly well known. You, you have commissions, commissions all over the place. Yes, all over so many commissions, 150 commissions. It's about almost triple what Los Angeles has, and they have 10x the population. A lot more people, right? So is. Artificial intelligence or just technology a way to solve that, or is it simply the way to govern? It absolutely need it can be helpful, uh, to us. I, I, you know, it should not displace people. It should help the people of our city, the workers at, at City Hall and all of our departments deliver more effective services for the people of San Francisco. should free up our employees to do better work for the 840,000 residents and you know sometimes technology works incredibly well and sometimes it doesn't and and and so we've got to be thoughtful about how we implement new, new, new things and I know that's what. The private sector is always looking at and we at the government sector, we tend to be slow to change. I'm going to push forward because what we have been doing in San Francisco in terms of delivery has not always been the most efficient or effective, and we're going to try to lean in on it and use technology. Uh, it's right here, so let's test it out. Fair to say move fast and break stuff a little bit for I, I don't think we should be breaking things, uh, uh, in government, um, but we should absolutely use tools that are well, uh, you know, well regarded, uh, um, that have been tested, um, always with safety, always with, uh, uh, that in mind and regulations and guardrails, but, but, um, I absolutely think we have some amazing. Innovators and entrepreneurs and companies here in San Francisco, we should be using the products. So final question, um, you're almost 1 year in. I asked you backstage if you felt like you were just going, going, going, and, and you, um, and if you ever got tired, and you said no just as you did at the top of this conversation, and then you said I've got 3 years left. It's a ticking clock for you. What are your priorities for the next 3? I think, um, I, I wanna leave San Francisco, uh, in a place where people around the world are like, oh my God, I have to go visit San Francisco. You know, we're, we're here. I know we're at an AI conference, but our number one industry is not technology, it's tourism, uh, and we have been the envy of the world for generations and, you know, hundreds of years now. Like we got away from that, uh, for a few years. We took it for granted as the elected class in the city. Like I'd never been in politics before. I was never, you know, I came from, uh, another field and, and, and people, you know, I think there was probably 4 or 5 people that thought I had a shot at winning and, uh, and so what I, how many were the family, yeah, probably 4 or 5, yeah, and, and so I think, um, what we are trying to create. Is is once again, uh, San Francisco being a world class city that is the envy of the world, and I think we're closer to that than most people realize considering where we were, you know, a year or two ago. Uh, this city is a city on the rise. Uh, we have. We are the innovative ecosystem of the world. That's why this conference is here and and and we also have the most incredible arts and culture scene. We have the greatest restaurants, I believe, on the planet here. Just great a great creative class and we happen to have some pretty unbelievable AI companies here as well so we have it all, uh, we just have to go tell the, tell that San Francisco story to the world and I hope everybody out there is paying attention and come see us here in San Francisco. There you go, Mayor Laurie. Thank you for joining us. Thank you very much. So may I thank you very much. Thank you.

As AI grows more capable, the risks become less obvious. The future of governance will depend on leaders’ ability to manage systems that learn continuously, behave unpredictably, and increasingly intersect with evolving regulation. Building the right metrics and oversight structures will be essential to surfacing inequities, strengthening training pipelines, and ensuring organizations stay accountable. How can we position governance and compliance as true drivers of enterprise value? 

The Next Phase of Governance

AI may generate possibilities, but humans generate progress. The real transformation happens when technology amplifies, rather than automates the creative process. In this discussion, executives share how they’re cultivating originality, driving impact, and sustaining growth in an age when machines can make almost anything except meaning. 

The Business of Creativity: Where Innovation Meets Impact

Your new digital workforce can be your greatest asset or your biggest liability. In the agentic era, you can't have speed without trust. This session will explore strategies for designing, deploying, and securing agents to ensure observability and control from day one. Discover how senior executives are securing and governing AI agents, while turning security into a strategic enabler for innovation and scale. 

AI at Velocity: Securing the Agentic Enterprise

Can one platform unlock the true business value of AI? Databricks CEO Ali Ghodsi believes so. He discusses the company’s ascent to its mammoth $100 billion valuation, the urgent need for enterprises to turn information into insight, and what the future holds for one of the largest companies in the private markets.

The $100 Billion Data Question

Autonomous and chaotic, wasn't that great? Don't you love the juxtaposition? Ali, thank you for joining us. Of course, my pleasure. All right, we've got a lot to rip through, so let's do it, my friend. Um, Data bricks started as a way, if correct me if I'm wrong, uh, as a way to write code and run it on a distributed computer. Now it's a unified analytics platform. You've got data warehouses, data lakes, data zeppelins. I don't know, um, you let engineering teams build and deploy their machine learning and artificial intelligence apps, and you prep their data for it. Is that a good assessment of what data bricks is? Yeah, we got all the buzzwords in there. We got them all. Yeah, did I say big data yet? OK, good, um, so what do companies need more? What do your customers need more? Do they need the clean data side of, of that offering, or do they need the custom tools? They need security and governance. That's like the number one thing that's holding the big enterprises. Like if your company is more than 10 years old, the big thing holding you back is that you can't actually do anything because you're so worried about getting hacked. Who's gonna get access to the data? Is the AI gonna suck up the data and memorize it and leak it? You know, there's lawyers now taking over. It's like AI lawyers now in every organization. That's everywhere. So when you say AI lawyers, lawyers who specialize in AI or actual artificial intelligence, yeah, yeah, no, actually they specialize in artificial intelligence, and they like tell you that, hey, here's the regulations. Here's which models we can use. Here's what's our policy. Here's what company contracts with our customers say, so we're not actually allowed to use this model. But if we do, we have to use it in this way. We have to have guard, and they enter all the and everything slows down in all the organizations, right? They enter the rooms and that's, so how do you do governance? How do you, and there's regulation popping up around AI all over the planet, right. And you know, so that's, so for you it's, it's corporate inertia, yeah, the biggest blocker, yeah, absolutely. And for those companies that's what I mean, you know, they're like, hey, this disruption with AI is coming, but you know we're being slowed down. Can you help me speed up? What do I need to speed up? Yeah, yeah. Now a lot of companies are looking at AI agents as a way to do that. You guys have expanded into that area, right? It's Agent Bricks, Agent Bricks. And so what are you seeing in that arena? We've been talking about it quite a bit in the last. 36 hours or so, um, what are, what are, uh, your customers trying to do with agents, you know, where are we, where are we now and where are we going? Well, let me just segue into it from what I was just saying. Your agent is just gonna go access workday and have access to HR data where salary information and so on is. How do you like lock that down? How do you make sure like suddenly it's not leaking information or going into wrong? Systems and accessing stuff like just this authentication security who's allowed to do what is uh is a huge ordeal inside any organization so we do that that's unity catalog. But what are people doing with the agents um there are lots of cool, uh, use cases, you know. I think the marketing stack is being marketing inside the organization is completely being disrupted. Um, so take 7-Eleven is a customer of ours, and what they do is they just use these agents to prepare all the material that before was taking huge amounts of time to prepare to target different groups, you know, what should I show you? What are you interested in? What's your campaign? That's now being largely automated and the content is being auto generated. The groups are being segmented, so that marketing stack is, you know, it's, it's, it's a big and actually at theater as well, a big disruption coming to marketing itself, a little bit of dog fooding too. We're gonna actually talk about that in one of the lunch round round round tables today, um, so. Everything in this category is about fast growth. Your company, uh, the technologies that you're dealing with, the, the, the com the company is trying to change rapidly, right? Um, are you, for your business, concerned about the rising costs of AI eroding your margins? No, the way I think about it is that, uh, you know, it's just so valuable, these use cases, and you know we can so easily see the ROI for our customers that I think actually we've seen that we can actually charge quite a bit for this. I think it depends on which layer of the stack you're in. I think the foundation model layer seems very competitive and the margins are very low or if you're in the business of like renting GPUs and reselling them, you know, that's hyper competitive, uh, but I think this, the layer on top where you're actually having agents. Do useful work, um, I don't think actually it's gonna be a big problem with revenue and monetization. I think the big problem is one, security we already talked about that. The second is quality, like, you know, are the agents good enough to do work, uh, inside of these organizations? And I think, you know, generally speaking in the industry, the answer is right now kind of no, even though like I mentioned one disruption with marketing. There are other disruptions I'm happy to talk about, but, uh, generally speaking, it's not the case that inside of our companies we have lots of co-worker agents running around that we work with all day long. You probably have like 0 coworkers that are agents. Anybody have any coworkers who are agents here? Raise your hand. 00, we got a hand. We got a couple, not like human agents, but we're talking about AI agents here. Yeah, alright, we got a couple of hands. This isn't, I mean, brainstorm AI. Hey, I'm saying the disruption is happening, yeah, yeah, in different verticals. Yeah, um, so let's talk more squarely about your company. So you, you have raised, as Ali said, uh, in introducing us, you've raised billions of dollars from Amazon, Microsoft, Nvidia. I've heard of these companies, um. Your evaluation right now stands at about $100 billion you know, for just to like ground that in reality, um, ExxonMobil, you know, that company is worth about half that in the public markets right now. So just like that's how big that number is, um, and they're number 8 on the Fortune 500, by the way, like big, um, be number one forever, right, right, that's right, um, you've read your Fortune magazines, cheers, um. You're flirting with an IPO. Why do you need an IPO and when? Yeah, I mean, um, why do we need, uh, uh, an IPO? It's a good question. I mean, it's like, you know, I would just say access to more capital, so I'm not sure we need that much capital. There's just much more access to capital in public markets. Uh, the second thing I would say is that. It is. I do think for American companies like in the United States, uh, we're global, so we sell in all the, all over the world, uh, there is a marketing boost to companies that are public. There's like free marketing every 90 days that they get, uh, with earnings call and everything, um, so that would be another thing, um, yeah, uh, and then, you know, maybe a little bit easier liquidity for employees. And investors, um, but it's close. It's not like, wow, there's like huge advantages and pros and cons. That's why we're not super religious about it. We will go public at some point, but to us it's like not a really big deal. I know everybody's super curious and interested about it, but for us it's like, uh, you know, it's close enough that, you know, we might do it, might not do it right now, so next year maybe. Depends actually, I'll be honest, the way I think about it is that, yeah, the way I think about it is that if, if there is a bubble bursting, which there might not be, uh, then it would be great to be private because we can continue investing because one of the biggest advantages Databs had was that in 2021, most of my peers, CEOs, they're like, we gotta IPO our company, we're gonna go public, and I told them like, hey, maybe we should hold off. It's like this whole situation is kind of out of whack, and those companies went public in 2021 and in 2022. You know, inflation, interest rates, are you suggesting the markets have been volatile? Yeah, exactly. And what happened is that they were like in cost cutting mode, layoffs, and that year we hired thousands of people and we continued investing, and now we have, you know, doubled the growth rate of some of those companies. Uh, so I think it's, you know, just having like a long term vision and just following that, I think AI is gonna be super useful to society. So I just wanna do that over long periods of time rather than having to follow like market hype. Yeah, I think, I think there's quite a few folks in the room who might agree with you on that one. so let's talk about some of the partnerships you've made in the interim. Um, you did a deal with Thomas at Google Cloud in June, who was here yesterday. Um, you did a deal with Brad at OpenAI who was here earlier this morning in September. Um, you know, I asked, uh, Corewave CEO Mike earlier today about the circular, uh, AI economy just briefly because we're running out of time. I will ask you the same thing. He believes it's a fundamentally flawed thing. It's not the right way to describe it. You're doing those deals yourself. How do you feel about that term? Well, let's separate something. There's not like, literally the, there's only like the arrow just goes in one direction. There's no, there's no arrow back. Like we're like, so there's a separate. It's like there's, there's deals a linear. There's, it's there's deals out there that are like, you know, this vendor invests, let's say 100 billion in this company and that vendor buys, say, products. From the same vendor back for 100 billion, there's nothing like that going on here. What Databix is doing is basically saying that, uh, you know, if you commit to Dabi as a customer and you, we already have 70% of Fortune 500 doing that, you get open AI, anthropic, and Gemini models in the platform out of the box because there's such a tight race between these, you know, now this one is better, or now this one is better at coding. No, this one is a better chatbot. No, this one has agents literally every week. Yeah, every week, yes, so we're just saying don't worry about it. Uh, commit to us and we make sure that you're covered and you have all of them. You don't need to even worry about getting contracts or anything, and we have all of those. We also have all the open source models, including even the Chinese ones. So thanks to you, you need to catalog our governance layer. Some organizations decide to lock that down and say we don't want to have a Chinese models, so then they can do that. So that's the, that's it's just one direction. There is no, like, you know, in return we'll buy. You know, tons of tons of something else from them or something. There's no, no, no such thing going on, understood. Um, I wanted, yeah, I can, yeah, go ahead. I can bite on the, on the, on the circular thing though if you want, please, please, by the way, just to make it interesting, um, I, I think that those circular things are just getting started now, so the inflation only like the, you know, the blowing air into the bubble is gonna start now. Is there a bubble in your opinion? I, I think there are, there is a huge bubble in many, many portions of the market. There's like, well, I mean the companies that are worth, you know, billions of dollars with zero revenue, that's clearly a bubble, right? And it's like insane. Like if you talk to VCs and investors, I mean they all know it. They're like this is like maybe I should just go on a break for like 6 months and come back and uh it'll be like really financially good for me. That's what the VCs will tell you. Uh, so there's definitely a bubble there, but I think this, the circular. Uh, aspect of it is just gonna this inflation of that is starting now, so I think like 12 months from now it'll be much, much, much, much worse. And I actually think it's sound that the markets are kind of reacting to that, right? They're wobbly and they're not growing. I think that's good because it's kind of telling the CEOs out there that, hey, maybe, you know, you should, uh, take a step back. Interesting. I just heard yesterday somebody saying in a year we're not even gonna be talking about AI bubble, so we'll see. We'll be back here in a year. Don't worry, um, OK, so, uh, uh, one more question and then I would like to come to the audience for them, um. You've said data bricks in the past data bricks could be a $1 trillion company. Um, we only got the first one of those, uh, a few, a few years ago and it was Apple in 2018, something like that, um, and that's the public markets, right? I mean what do you need to do to build, buy, partner. To get there, yeah, so you described very well in the beginning with all those words that you threw out of what databix does. I think the word's good. Yeah, you're very good with that, but you know there's a couple of new vectors that we're getting into that are completely new for us. One is transactional database market. This is the The traditional Oracle market, uh, that market has been the same for 40 years. So for us, the lake-based offering that we, we, we launched just this summer, that's a very, very big bet for us. Um, why is that interesting? Because for the first time we're seeing over 80% of the databases that are being launched on data bricks are not being launched by humans but by AI agents. This is actual stats. But we might ask them, well, what is this? Like we're already. Agents, it's people doing vibe coding, so lots of people are writing software. They're vibe coding stuff. Well, all that software needs a database behind the scenes, and the software prefers to use Lakebase. It's really built for that, so it's best at that. So that's just a huge growth vector for us, I think. If we just did that, we could maybe get all the way to 1 trillion. The second one that's really important for us is Agent Bricks. Yes, and Agent Bricks is important because Again, the real opportunity is, you know, it's a commodity now to have AI that has general knowledge. Ask it any question you want, as you said, the race is like going, it's a commodity. You can get free open source models, but something business specific is, but it's very elusive to get AI that really rocks and understands. That proprietary data that's inside the enterprise. So we can actually unlock the value of that and start automating more and more and more and augmenting the work that's happening inside the enterprise. So I give you like the 7-Eleven example with marketing. Here's another interesting one like in healthcare, Merck built a model model, AI model called Teddy. Uh, and it's a generative AI model, so it's, you know, stands for transformer enabled drug discovery, and this model is really have amazing understanding of cell behavior, and it can really differentiate between causal cells and reactive cells. So it has the opportunity. There's a lot of value there, yeah, it can really revolutionize drug discovery. uh, so that's another one. Another favorite of mine is RBC, uh, Royal Bank of Canada. Uh, this use case is equity research analysts. They do all this. Research on like oh there's an earnings call here there's an earnings call there, but what they've basically done is they now with agent bricks they're building these agents that get the earnings calls, get all the information that's out there about that company, everything that's going on, put together the research report and uh just augment significantly, get, you know, many days' worth of work down to like again yeah if we can just keep unlocking uh sort of uh agentic sort of use cases like that, I think it's going to be that that that. Can take us all back. That's your ticket to the 3rd, OK, to get the trifecta is also to have apps on top of this. All right, now you have apps that are vibe coded with the database lakebase and with agents. Um, OK, those are 3 new vectors for us. All right, we've got a plan. OK. There's a question right back here. Yes, please, your name and who you're with briefly. Steven W Preda, CEO of Alpha, it's a govern, AI governance intelligence company. I was at the summit in June's great job with, uh, Jamie Dimond, um, but you're hitting on a point and you kind of discussed it a bit with Jamie at the time. The governance issue, especially in the boardroom. And so you really see this kind of push pull between focus on all the risk, but they're really not talking about the growth opportunities. How are you actually seeing boards really kind of challenge the governance issues because you're obviously very important to that whole conversation. Yeah, that's a great question. I mean, I think that, you know, uh, mostly it's actually problematic because the boards are just pounding the table and saying, hey, like, where is the AI revolution? Like, you know, you're spending all this money, where are the results? And they're kind of frustrated and then the results are not coming fast enough. The reason the results are not coming fast enough is that there is a legal and a security aspect to this, uh, you know, and that, so that's what we talked about in the catalog where, where we work on that and there's new legislation coming and all of that. But then separate, even if you can figure those things out. The data architecture inside most organizations is an absolute mess because they've been buying software for the last 30, 40 years from different vendors and just piled it on. A lot of the data is siloed on premises in data centers still it's not in the cloud, and then in the cloud it's in different clouds. So just getting, you know, their ducks in a row there is also, you know, problematic. So most organizations are frustrated. There's a big gap between the CEO and the leadership and the board who's saying let's go, go, go, go. Why are we like AI revolution is already here. And then the reality on the ground, which is like, well, you know, give me another year and give me more head count and I need to spend way more to even get there, uh, so that's what we have to resolve and there are ways to actually solve this, um, you know, 11 of, one of the things I like to say is like, you know, I think the process change management with humans is actually the biggest problem, not the tech, of course, yeah, so I think the biggest issue right now is that now inside of the organizations there's like. Many people who say, no, I should lead data and AI strategy for this company. I'm like those guys over there, they don't understand anything. So there's like this tussle between different people who are fighting over, you know, becoming that AI person, and that's also making things worse. So like my advice to leaders there is, can you like pick one person for your company? Don't have a three-headed, you know, monkey. Doing that, it's always the humans. Thank you for your question, Ali. We're out of time. Thank you so much for joining us at Fortune Brainstorm AI. Thank you, sir. Thank you.

Autonomous drones and pilots are flying places humans cannot go, while national security increasingly depends on data as much as force. Two prominent founders of defense tech unicorns discuss how Silicon Valley and the Pentagon are rewriting rules in a rapidly evolving (and unstable) geopolitical landscape. Where will AI redefine deterrence, surveillance, and battlefield autonomy fastest? Where is tech making progress on the challenges that the U.S. defense establishment faces––and what remains elusive?

Defense Disrupted

Hello again. Tara, Gary, thank you so much for being here. It's great to be here. Thanks for having us. So let's start really big picture, um, what are the biggest challenges of selling to the largest bureaucracy in the world? I don't think we have enough time for that question, Ali. Uh, I'm happy to start, although, Gary, I bet you have plenty of war. I do it, you can do it. I believe, um, it's speed and it is the restraints that are placed on federal government buyers and the ability to take great ideas, smart people who work in government who are trying to do good things and. Help them break through the bureaucracy and access funding. You know, we spend at least 6 months out of every year under a continuing resolution, meaning that year's budget hasn't even been granted yet, so it really compresses the ability of time to be able to drive this change and get that deal closed. Yeah, and I think about it internationally. I don't think about just the US government. I think about global governments and how we can better serve warfighters around the world. And the one thing that's really obvious and the thing that gives us passion every day is like how do we put better tools in the hands of warfighters and really make a difference. And the good thing is I think there's an understanding that it has to get better, and I think there's a lot of well-intended motivated people to try to make it better, but. As Tara described, like we're at the beginning of that journey to see radical improvements, and those radical improvements are absolutely required. And you think about everything you guys have heard here, the rate of change that's happening, which is super exciting, you need that rate of change to actually land in governments around the world, and they need to see the positive effects of that technology change. We're actually just coming off of one of the biggest defense conferences of the year where there was a lot of messaging around this. The Pentagon is saying things like it's time to move to wartime footing, for example, and really accelerating rates of adoption. How much of that is real and how much of it is theater? I'll start, um. I think there's very good aspiration to do something fundamentally different, but we're, it hasn't flowed all the way down into the bureaucracy. The one place that I would say it's happening every single day is in the Ukraine. So at SHIELD we have employees deployed in the Ukraine working alongside the Ukrainians leveraging our technology to get better outcomes against the Russians. And what we see there is this total agility where something works one day, you get feedback, engineers work overnight, we get them fixed, and it's back into the battlefield. And I think that agility, I think everyone's watching it around the globe. Everyone's watching really closely what's going on in Ukraine, and it's that agility and ability has constantly changed. It has really been one of the most important aspects of Ukrainian defense, and I think we as a nation, the United States, need to be better at that. Agility because that's the way the next conflict is going to get fought. I think building on those great points, a big part of what Secretary Hegseth's message is and what the US national security apparatus is trying to do from a move to wartime footing push is to really catalyze the American industrial base and drive production because much like we've seen in Ukraine, the reality of A future war that the United States is likely to enter is going to deplete stockpiles that are already anemic today because of what we are supplying to allies and partners around the world and rightly so because it protects US national security interests as well. Unfortunately those stockpiles are low enough and the United States has outsourced so much manufacturing capacity. At this point that the amount of time it will take to build the munitions and weapons systems that the United States needs is just much, much too long. That's why we have to get on wartime footing now, especially through the lens of the industrial. How long is too long? What are we talking about here? I mean, a lot of war games will show that in a notional war with China, the United States is out of munitions in 7 days. Whether you think 7 days is realistic or you think well maybe the beginning of the war will play out differently, you're still operating in things like days, weeks, or months, and the reality is when you look at Chinese war doctrine and the Chinese Communist Party's. Significant amount of time and attention spent studying, going to war with the United States. They are planning for a very protracted conflict and would be happy to draw that fight out to bleed American stockpiles dry because they aren't missing the economic piece of this puzzle. I'd be curious from the audience who here would say they're sort of familiar with the state of defense tech. Raise your hand. That's not that many actually, um, so I think it's maybe important to say, um, you know, I'll start with you Gary since you brought up Ukraine. I know SHIELD has been deployed in Ukraine had technology has been deployed in Ukraine. Um, what have you learned on the ground there and how much did the war in Ukraine change the state of defense tech? I think that the conflict there informed the way conflicts can be fought. There's a lot to be learned, but there's, I think, a conflict with the US would be different. But one, it's dominated by autonomous systems, whether it's one-way attack drones or more sophisticated autonomous systems. You're just not going to put. Warfighters at risk if you don't have to, and both the Ukrainians have learned that the Russians have learned that the speed of change is incredibly high because certain tactics work and then you have to regroup quickly. So one of the very simple examples is GPS is basically denied 100% of the time and comms are denied, meaning that if whatever weapons system or aircraft you had that was dependent upon comms or GPS. It's just not going to work in that conflict and so you learn and that could be life or death and that could be life or death, and it could be life or death for sure. And so being able to respond to the actual conditions on the battlefield have been absolutely critical and I think through those experiences for people like Shield, if you think about it from a product development perspective, you get a whole lot smarter about how do you drive. The innovation cycle and the iteration cycle that's absolutely necessary to ensure that you're delivering something to deliver mission outcomes and so I think Ukraine has really. Shown that you can be agile, that you can use lower cost kinds of lower cost kinds of systems, and that you can iterate quickly and get real outcomes, and I think that's that's so important in the world that we're living in today that is under so much change. And Tara, how has Ukraine affected defense tech from your perspective and particularly how you think about readiness? Well, I think that it has shown the urgency that we all feel in sort of an abstract way in terms of just how quickly those learning cycles have to happen. So today the. The United States is in a readiness crisis. Our weapons systems and military platforms have historically low operational availability, which basically means if we need to go to war, half the fleets sitting in depot or at dock. In what we've seen in Ukraine is that adaptation is probably the most important skill and so what we are doing as a company at Gvini is using data and software and AI in order to try and bend that curve on readiness. And I think to Gary's point, it's both making sure that you have the right set of systems, many attritable lower cost systems, and addressing that cost curve. Today in the United States we'll use a $20 million missile to shoot down a $2 million drone overseas. That doesn't work and it shouldn't work for US taxpayers. So how do we increase military readiness and also start to think about what the conditions of the future of warfare will look like and ideally this we sort of take for granted in the United States, but ideally do it in a way that this country can afford for the long term. And when you think about AI since we're at Brainstorm AI, um, how fast is AI being adopted in the military right now? Over the past, I would say 12 months, much, much, much faster than we've ever seen before. What's spurring that on? I think a lot of it is a combination of a maturity of technology and frankly a presidential administration that wants to work with tech and once you open the doors to tech industry and of course credit to the tech companies. For saying we want to support the defense mission and US national security, that's what happened first is the the company showed up, then the investors moved in and said, OK, we're in this too, we'll give you the capital to be able to develop your products with a national security or military application. And for a while the Pentagon was still like, yeah, we're not sure what to do with you, and we don't know if we can trust you. Aren't half of you gonna go out of business and, uh, and what we've seen over the real side eye there, yeah, exactly, and the reality is actually 95% of companies will go out of business, but the 5% are gonna change the world, and I think that's. Exactly the appetite that this administration has, which is let's find the 5% that are gonna drive this significant change and uh try and reduce risk in acquisition. I'm sorry, accept more risk in acquisition in order to reduce risk operationally. What about you? And, and I agree with Tara in that. It feels a lot better, but we are still, we have so much more to be done because there is so much opportunity, and I, I look, one of the things I'm most excited about is. The fact that AI will fundamentally transform the aerospace and defense industry, it, it's going to look different in 20 years than just radically different than it does today. You're going to have lower cost systems, AI led, software-led, um, not these super expensive. Incredibly elaborate systems that just get shut down and so I, I think there's a revolution happening and we're at the very beginning of that journey, um, as Sarah said I think the administration is really open to this, but we're, we're 9 months into that so a lot more to, a lot more to come on that front and we've got to move faster. We're going to open the floor to questions, um, but I am curious, uh, you know, the Pentagon is famous for moving slow. Silicon Valley is famous for the phrase move fast and break things. What is the biggest cultural clash between tech. And the Defense Department. We get to wear jeans. No, I, I think, I think, um, risk tolerance is the biggest clash and um. I mean, I'll use a company that isn't here on stage, but you know I certainly think highly of as an example. Look at the recent coverage in the Wall Street Journal of ARR and highlighting the failures of that company as sort of a way to put a wet blanket, what I would characterize as a way to put a wet blanket on what's happening in defense tech and what I think that story actually. Highlights is the reality that in technology when you are starting to really do innovative things that are brand new you're gonna see a lot of failures and that's totally fine so long as you're learning from them rapidly and adapting your own businesses and making products better and in the national security sphere especially around the US military. Risk tolerance is exceptionally low and frankly for quite good reasons you mentioned it earlier, Ali. This, this is all life and death and so I think both sides are genuinely trying to figure out what's the right balance of that when the money we're throwing at this innovation and experimentation is taxpayer money, so we have to be responsible about it and these are. Serious systems with very serious outcomes, we have to be able to rely on their lethality on the battlefield. And yet if you only ever play in the safe space and you don't take some risks and you don't embrace some of that failure, we're going to continue falling farther behind. And so that's the biggest clash that I see. Yeah, and I can give you a slightly different perspective, but building on that. So one of the things I'm most excited about, so at SHIELD we're building a next generation autonomous warfighter. So what is that? That's what does that mean? That's a plane that takes up vertically, lands vertically, flies long distance, 2000 nautical miles. OK, carry's munitions, like they would be the same. Same that you would see roughly on an F-35, 1/10 the cost, lifetime cost of an F-35. But what's exciting about it is if you think about the people that had to build the F-35, their number one thing is about not killing the pilot, right? Making sure that in all circumstances you're not going to kill the pilot. In our world, we're building an autonomous aircraft. We can go hard and crash. And try things that you wouldn't be able to try if you're engineering with a human pilot because you're going to be safety is going to be your number one priority in this world where it is autonomous, you can cycle fast. So think about the agility mindset in engineering and being able to try things, experiment, and, and iterate fast. What's also different about this next generation of young companies. Is we're raising private capital so it's not on the, it's not on the taxpayers' back that we crashed one day. It's, we have to be, we have to be prudent to our investors we have to be prudent and thoughtful to our investors, but that speed of iteration is what ultimately creates value and shareholder value for investors and you can be aligned on that, but it's very different if you're using government money and poor tax papers watching some auger into the ground. But I love this. Fast agility, test fast, go fast, and drive innovation that looks fundamentally different than what we've seen because you're not focused on, oh my God, we've got to protect the pilot. Do we have a question from the floor? We have time for exactly one. If not, I have one final question. Yes, please, name and company, and you're bringing us home. Hi, my name is Margaret Tong, and I'm uh building an AI agent to automate planning for aviation programs. So as someone from aviation background, I'm just curious about, um, like you mentioned, safety and risk is a very big um concern in our industries. So like, what kind of regulation, um, restrictions or challenges are you facing right now? And do you see that, um, As something that is like overwhelming you, like you said, like it's pulling us back from like letting AI to do more things for us, or actually it's something that we do need those safety guardrails to make sure that, for example, in terms of like cybersecurities, when we get hacked or electricity just gets stopped, like hacked by foreign countries, like um how do we make sure the safety is still in place when we have AI to do a lot for us? Sure, I think. So as we think about the broad profile of delivering final autonomous aircraft, we have to think through all aspects of safety, um, including cyber so what is that? Software development cycle, how do we ensure that we haven't, that we're tamperproof, that there hasn't been something that has, um, taken over the code, and one of the things that people think about is, oh, somebody's going to press a button and the drones will fall from the sky. It doesn't really work like that, um, and so, but that is a concern, and I think for us the burden is on us to demonstrate that we have safety throughout the process, whether it be cyber safety or flight safety. But we use a very strong path of iteration to test and ensure that we can deliver on those qualities and then we certify against a set of um kind of classic certifications that are absolutely required to give our customers comfort that we can deliver on those values so it's a very high priority it registers. Um, as probably the most important thing we do, but we don't operate with the mindset of killing human pilot, which has held back a lot of the speed that has traditionally been needed in this industry. We're out of time, but one word for what the future of war looks like, Gary. Autonomous Chaotic. Thank you guys so much.

What happens when every customer comes with a personal shopper? Agentic browsers that understand intent and AI-powered marketplaces that translate taste into choices are drastically changing the way we buy goods and services. When consumer discovery, decision, and delight can be intelligently automated,what's left for the retailer?

A New Kind of One-Stop Shop

CoreWeave’s IPO wasone of the defining moments for AI in 2025. CEO Michael Intrator joins us to reflect on the company’s first year as a public company, share insights on the evolving AI infrastructure landscape, and discuss how AI is reshaping the future of the cloud.

The Future of AI Infrastructure

Good morning. Mike, thanks for being with us. Absolutely, um. So, give me, give me a second because I want to explain to everybody what Core Weave does. We've got a live stream on, you know, it's a little complicated. So Core Weave, and please correct me at any point if I get it wrong, Coreweave creates and manages data centers, sells their computing capacity to tech companies. Um, in, in my view, if, if there's a company at the center of the AI boom, it's you. Um, Nvidia owns a 7% stake. That's another company considered to be at the center of the AI boom. You don't own the facilities. You're trying to rapidly diversify your customer base. You have billions in short-term debt and billions more in long-term debt. It is a high wire act. I do not envy your, your, uh, you, sir. And looking at your stock price since your March IPO, it has been seesawing, and to my eyes, I'm seeing the push pull of these guys are the future and these guys are going to be out of business. It just feels like everybody is fighting over that. Do you feel misunderstood? Um I think uh we gotta go, go through that introduction of a few things um so look, um. Core weave. Delivers intelligence. To consumers of compute that require that product to build the most cutting edge, uh, artificial intelligence or even more broadly that require parallelized computing we do that by building. A beautiful software solution to allow for the most performing configuration of GPUs to the market. That is a function of a software layer that we have created um that is singular. It is the best solution in the world for uh providing access to compute. For those consumers, um, we are well known for that, um, and that's your focus. You're not for everybody, you're for this higher end compute. We are for when we, when we started the company, really we started with a blank sheet of paper and say how do you purpose build a software stack? How do you purpose build a company to be able to deliver the best product for consuming compute that is occurring. Um, in a parallellyzed fashion, and that, that is what the company does, um. We do, uh, build our own data centers. We also co-locate our, uh, uh, compute within third party data centers. Um, we are one of the largest, um, um, resources for this type of computational power for the supercomputers that are able to do this in the world, um, so. You know, the, I, I, I really think that, you know, we're not a company that rents GPUs. We're a company that provides a, a, a, a really wonderful solution and you can see us as out there, we're moving up the, the software stack, making a more robust environment with acquisitions like. Weights and biases and you know open pipe and you know like like those are those are acquisitions but in addition to that there's this incredible kind of organic build out of the software that we provide so um that's the way I think of this and and and I think it's a it's a more accurate, um, um. Uh, representation, I probably do it more often than you do. Um, indeed you do. Uh, I believe it's part of the job and, and that, and that I will therefore come back to it, you know, in the run up to the IPO, you obviously had to go on your roadshow, you know, you, you've, you've been out in the market since. Do you, I'll ask you again, do you feel misunderstood? Are you, did, did, does everybody else understand what you just said? So look, um. Let, let, let's, you, you, you're, we're building a company that is challenging. One of the most stable comp uh. Businesses that have existed, right, the cloud business has had these 3 massive players and then you go through this kind of changeover of fundamental technology. You go from sequential computing to parallellyzed computing, um, you have the advent of artificial intelligence. All of these things kind of occur at the same time. The GPU makes these massive strides forward so that you can, uh, uh, engage in computational power at an order of magnitude that the world had never even contemplated before. And all that kind of comes together just as Core Weave rolls out its business, so. Do I feel misunderstood? I feel like it's incumbent upon Core Weave to introduce a new business model on how the cloud is going to be built and run. Into the future and that's what we're doing, um, you know, when you talk about our stock seesawing, um, I, I think people are myopic a lot of times. Yes, it is seesawing. We came out into one of the most challenging environments right around, uh, Liberation Day, um, and in spite of the incredible headwinds we're able to launch a successful IPO. We were, uh, the, the IPO price was 40, the stock is trading at 90. Um, so yeah, it's been incredibly successful and I think more people are beginning to understand it, but when you introduce a new model, when you introduce a new way of doing business, when you, um, disrupt what has been a static, uh, environment, it's gonna take some people some time, um, and I had said that, you know, when I look back in history, um, of the company, it took us. A year. With, with a company, with, with an investor like Fidelity, we had to introduce what we're gonna be doing, we had to explain it, we had to come back and answer questions. It took us a full year before they were like, oh, I get it. And so, look, we've been public for 8 months, um, I couldn't be prouder of what the company has accomplished, um, you know, I feel like. We are leading the innovation within the software stack um in terms of providing uh the environment within which so much of the real cutting edge uh uh artificial intelligence is is occurring uh we have been uh innovative around. The physical components of building this infrastructure we're the first ones out there kind of looking to repurpose power whether it's from crypto or other use cases to access and build really really large scale infrastructure so that that the supercomputing clusters can be built. And uh we are uh the leading innovator around how you go about. Financing this infrastructure whether it was from the IPO to even yesterday when we did, you know, a $22 billion convertible, um, note that was upsized to 2.25%, it was, you know, 5X oversubscribed, you know, it really gives you, uh, insight and understanding to how. Many people from different sides of the balance sheet, both the equity side and the debt side, are interested in joining the journey of Core Weave and, uh, participating in the cap stack in different ways, and you know, I, I, I just think that like that represents. Core, we've once again tapping into, uh, new, deeper, more liquid portions of the, of, of the cap stack to be able to, um, you know, uh, finance and continue to grow, uh, this business that is in such tremendous demand. Thank you for that, by the way, um, I, I wanna ask you another question, but um we do have a question down here so I'll come to you next if you don't mind, um. Let's talk about your customers. Um, everything that you've just articulated is classic AI. Everything's moving very quickly. Everything's changing very quickly. The, the, the, the speed of change is, is extraordinary. What are you seeing in terms of change in terms of what your customers are asking of you? So 8 months is an eternity in AI. Yeah, I do feel old. Uh, so, so look, let, let, let's start with the customers, right? So when, when we came out into the IPO, um, I love how I'm using like this 8 months as if it's some kind. Of like, uh, it's a journey, it's a Lord of the Rings journey, yes, so, so, uh, when we came out we were, uh, incredibly, uh, one of the knocks on us is that we were incredibly dependent upon Microsoft for 85% of our revenue, um, currently at IPO at IPO. Uh, it was a stated goal for our company to drive diversification within our client base, and now none of our customers represent in excess of 30% of our backlog. And so, you know, when, when you say what are our customers, um, looking for, it, it depends, right? There are certain. Um, givens, what are they looking for? The most highly performing infrastructure? What are they looking for? They're looking for incredible speed of the most cutting edge technology to market so that they can gain a relative advantage over the companies they're competing with so that they can deliver infrastructure to the clients that are demanding access to the compute so that they can run artificial intelligence within. Uh, uh, their ecosystem so that they can monetize the investment around, uh, artificial intelligence that they've been engaged in investing in for, for years, um, but there's also new clients that are that are emerging that are, um, looking to, you know, structure deals in creative ways so that they can, uh, have an opportunity to launch a new business, a business that didn't exist in. An environment where it's incredibly difficult for new companies to access compute at the size and scale that will allow them to release products so that they can compete in a world that's dominated by these these foundation models that are so massive and have access to such amazing capital and we've been really thoughtful around how you help that that ecosystem kind of continue to to uh. Um, grow and be resilient and, and scale, and we think that's really important. Yeah, understood. All right, let's go right here. Um, wait for the mic if you don't mind. Good to see you, your name and who you're with for everyone. Yeah, hi, uh, Pankaj, AI investor and early investor in Corbi. We appreciate your support. Um, so, um. You know, the, the market is there, you know, I believe in 5 to 7 trillion worth of data center over the next 5 years. Korvi team is super strong. Your strategy is right. So I think of K we as an execution play at this point. And, um, you know, there was a little bump in the last earnings report and the stock fell like crazy, right? So my question, um, Michael, is, From an execution perspective. What could go wrong if data center deployments get delayed, if energy is not there, if water is not there, if talent is not there, if there is a recession, etc. How do you, what, what keeps you up at night and how should we think about that? Stealing my questions. So please tell us about your nightmares. Yeah, yeah. Um, So look, um, anytime. The world builds. Infrastructure at high velocity. Um, there are shock loads to different parts of the um supply chain, um, that are. Difficult to think about. I, I'll give you an example. I was on a call, uh, yesterday, um, where, um, the, the, the person who was speaking with me was a CEO of a mining company and he was talking through the, um. The, the, the impact of the data center buildout on. Uh, supply chains on metals on the timing of building new capacity, and it was just wild for me because it was like, man, that's like two levels deeper than anyone who is looking at building companies like the one I'm building is really thinking about. We think, hey, you know, where's our damn transformer? We don't think about the. Damn transformers held up because you know there was an issue within, you know, copper or whatever it was and it's just, it goes down levels and levels and you know they're really far downstream and you know as you go further upstream you ultimately end with the bottles and everybody kind of I think has an idea maybe two levels but not the full stack of what's going on. It's fascinating, um, so, so look, um, you know. I, I think it's really important when you're, when you're looking at a growth business, business that's growing and scaling like we are, and you know, I, I don't know that there have ever been infrastructure companies that have gone through the type of scaling that you're seeing at Core Reef, so. If a data center Like from an accounting perspective, there's a day that you mark the, you know, the infrastructure going online, and if it misses by 1 week and crosses a quarter, you know. It's bedlam, but it's not really bedlam on the broader scope of we are out there building and scaling infrastructure every day, you know, um, we added 120 megawatts of, uh, of infrastructure last quarter. Um, we ended the quarter at 590 megawatts. We're going to end the year at 850+ megawatts of infrastructure. Like you gotta. Like, everyone is so incredibly zoomed in on every hiccup and glitch and this and, but like when you pull out a little bit, you gotta, and this is what I do, right? Like, like in running the company is I try to zoom out a little bit and get my arms around these macro trends, these like super cycles that are going on. So yeah, the data center was held up by a couple of weeks, OK. The demand from the most knowledgeable, most sophisticated, largest tech companies in the world is relentless. That's what the trend that matters to me, right, um, as we get larger, and I always refer to this as my denominator problem, right, like if you have Uh, 500 megawatts and you go to 850 megawatts and you missed by a week, it's a big number. If you're, and that's one data center that you miss on, right? If you have. 50,000 megawatts running and you missed by one data center, you guys won't even see the number, right? It'll just be lost in the noise so you're, you're, you just gotta zoom out a little bit and understand like these are macro super cycle trends that are occurring we are deeply ensconced in it we are able to access the information at the level that is picking up. Clients from across the entire space as they plug into our infrastructure and we know where the demand is coming from and we're building as quickly and effectively as possible. Mike, I, I, I have another question for you. Our time is rapidly dwindling, so I, I, um. I'm gonna leave aside the one around how you feel that uh the circular AI economy is a fundamentally fundamentally flawed premise as much as I'm interested in that, um, and I'm just going to simply ask you what do you feel is holding the AI economy back? You were talking about this relentlessness and the challenges of dealing with this absolutely insatiable demand. What is holding us back from being our best AI economy selves? I'm gonna go back to the uh oh you can't. No, no, no, it's just because I'm gonna tie it back to to to the conversation that I was having with uh with the uh the mining uh uh CEO and we're talking about mining a lot more than I anticipated. He, uh, he, he was, he was, you know, and he was just like, look, to, to, to do this, we need to work together as a group, and I said if I do that, I get accused of being in a circular economy like. Companies are trying to address a Violent change in supply demand you do that by working together. So that's all I'll say on the circular economy is like you do that by working together if that's circular, that's the incorrect way of looking at it. It's a lot of companies working to address an imbalance that is distorting, um, you know, the globe and, and this is how you do it and we've seen it before and he was asking to do it on the call, which I thought was fascinating, um, in terms of what's holding it back, you know, look, the, the, um. There are a tremendous number of brilliant entrepreneurs. There are an enormous number of, you know, brilliant data scientists. They are trying to access computes so that they can build these new products and deliver them to market. And, you know, right now there is a physical bottleneck associated with getting. Uh, the, the most performing compute into the hands. The most cutting edge players and, and that's really what's holding us up and we're working like hell across the entire space to ensure that we're going to be able to deliver as much of that infrastructure as possible, um, over the next several years that this demand continues to, to, to rise. Yeah, so not funding, not policy, just, just the physical bottleneck. I mean like. Funding is an issue. It's something we work on continually. Policy is an issue. It's something that you know we're all working on, but ultimately where the rubber hits the road, it's the ability to deliver compute into the hands of these scientists that are going to move the world forward. The reasons that you have challenges in. Delivering that compute is because of policy, is because of funding, is because of physical infrastructure, is because of energy, um, you know, and, and a lot of those themes are, are, are really talked about continuously, but you know, I mean, ultimately, you know, our job is to, to, to allow these scientists to build the world, um, uh, you know, in ways that, that we can't even imagine it. Mike, thank you for taking us into your dreams and nightmares of AI. A round of applause. Thank you, sir.

The explosive growth of data centers amid the AI boom has put incredible stress on infrastructure—especially the electrical grid. Tech giants have partnered directly with utility companies and billions have been committed to upgrades. Will capacity grow fast enough to keep prices from rising? Will policy keep pace? And is all of this enough to maintain AI's momentum? 

Scotty, We Need More Power

Calvin, hello. I, I feel like we ended the last one on a meditative note, and I'm noticing your mood beads there. So are you the calmest man in the power system right now? I'd like to think so, at least I'm thriving for that, but probably not, probably not. Well, look, I want to start with the fact that if we were doing a brainstorm with the head of a major utility company 10 years ago, we might have been talking about wind farms. And and solar panels and sort of happy consumers giving power back to the grid and our green cities and here we are in a panic mode about AI. So let's start. Has the dream died of of the first scenario of. Cleaner power and everything else. So first, let me say good morning and good morning to you. Thank you. I like the jacket. Um, yeah, well, thank you. I mean, I feel my son bought it and I feel like, uh, you know, Ralph Lauren should pay me to be wearing this big a thing on stage. So there you go. I, I would tell you, I would, I would tell you. We're not in a panic mode, 1st. 2nd, that dream has not died. We are a country that needs to continue to invest in all technology whether it's nuclear, wind, solar, gas. We need every electron. To make a difference we need that because we're in an unprecedented time where we're dealing with low growth and that load growth is being driven by technology data centers but also the onshoring of manufacturing, the electrification of everything, all of that is coming to a head and I say this all the time is that. I've been in the utility industry for about 25 years and in my entire career and probably the last 4 decades we have never had a moment of this amount of low growth and what a great opportunity we are in, but we need to seize it. And unfortunately for us as we get caught up in policy debates as to what is good, what is bad, what is, not just policy debates, you're actually election issues now in some of the areas that you're in, right? Was it the price of electricity is the new eggs, and that's not a good thing and. Um, so, so talk about that, even just the areas that you're in, incredible, that's both an area of incredible business growth but also affordability is acute in places like New Jersey. No, thank you for the question. So think about Exelon. We have the privilege to operate in some of the largest cities in our nation, Chicago. Uh, New Jersey, New Jersey, Baltimore, DC, Philadelphia, so you've got urban. Rural coastal and we have some of the largest uh disparity of incomes in the nation so affordability is key, which is exactly why it was an issue in New Jersey where but this is what you can't have progress if you don't invest, you know, so we want jobs we want economic development but don't do anything. It, it, it's not going to work. So what we are trying to do and we are partnering with our elected officials to say there is a thoughtful way of doing this and that goes right back to policy matters and I have 10.7 or almost 11 million customers to think about and I have to serve. The lowest income to the wealthiest, so you take DC as a microcosm. I have to serve the 2nd ward the same way I serve the 8th ward. Well, it's interesting because you actually do not own the underlying, um, power plants. You are the Constellation Energy that was spun off about 3 years ago. Now you want to get back in. So, so how do you want to get back in? Why is that important since we're talking in the context of Scotty, we need more power. Calvin, we need more power. Is that what it's going to take? You need to own this again. Well, let me just be very clear. I'm good. I don't want to get back into it, but if they're not going to build, someone needs to. What I always say is the independent power producers have no incentive to build anything new. Because they're maximizing their assets, which is fair for the current market conditions, and they were being when when low growth wasn't occurring they were getting pushed so now that growth is occurring they're able to maximize the dollar using New Jersey as an example. The average residential customer last year in New Jersey saw their bill go up $36. You know what, supply, the supply, the piece of that, it went up $38. I actually reduced their bill because of the reconciliation and blame and yet they blame me, and that is concern. So what we're doing is we're telling policymakers the warning lights are on. You're you're driving your car, the check engine light is on. You're like, I'm gonna keep pushing this. And no one's gonna pay attention until it breaks down and then everyone's gonna say, Calvin, why didn't you tell me? So what I'm doing is I'm telling you on that hottest day or that coldest day you might have a supply crunch and people are going to suffer. I'm telling you you have to fix it now. So that's why regulated generation is so talk about the talk you made this announcement yesterday with Next era. It's uh two states. It's intra utility red, blue. I mean, good for you that there's a man who's bringing us together here. Talk a little bit about why that matters. Is that something you think's critical? And again it's in the context of a conference where. The trajectory of AI and the trajectory of power are not necessarily on the same track. So Diane, let's go back to what we started with. You don't get data center growth, AI development, and so forth without building things, and that 765 KV line transmission system that we talked about announced yesterday with Nexttera is key because what it does, it's the highest voltage transmission system going across Pennsylvania and West Virginia, 220 miles, and what's so key. About that, it's helps with reliability, resiliency and affordability all at once. That that is the largest scale transmission system and it maximizes the output of the system. How real is the issue of of NIMB for those, you know, the not in my backyard phenomena, both with there's the data, you know, do you want a data center in your backyard and now do you want transmission lines coming across? Do you get much pushback these days just given the need or is it just every day? And what what I say to communities and what I say to data center developers what I say to my team members, you have to do it with the communities you can't the communities cannot feel you're doing it to them we're going to create jobs we're gonna create a tax base but you have to take their voice into consideration and those who do it well. Do that those who don't, they just show up and start building and it never goes well and what we started about a year ago really working with the communities on identifying the path of least resistance but you're always going to have some noise but you don't get to where we wanna go if we don't build something. Can you just anchor for us how critical a situation are we in? Because you know we're talking about accelerating growth next year, this whole conference. Did you really get the sense that 2026 is going to be the year where the curve goes up? Can we handle it in the current context that we're talking about with your six utilities? I think we can. Because we've always had, we've always shown and demonstrated our resilience in stepping up to the challenge each and every time. Now understanding I'm not a believer that the growth that we are talking about is going to be 10x or 15x, but I do believe we're going to have growth. I do believe we have developers that are playing states against one another and there's some double counting. And I also am optimistic because we do have the most resilient infrastructure in the world. We can maximize in the world. We do, we do, we do. When you, when you think, and I can speak directly, we will be, we will be putting as an industry $1.1 trillion of investment over the next five years, and you are chair of the Edison Electric Institute, I should say, so you've got the whole macro view of what's going on. I think Diane, you were trying to say I'm biased. Yes, no, no, no, I'm just saying. Our man in Washington in many contexts here, so yes, I, I am, and I, and I, I know my peers and we know our number one job is to provide a reliable and resilient system because it doesn't matter to you. More of you are working from home, you're educating your children from home, and if the lights don't come on, all the other stuff that we talk about, so you don't think about electricity until your rates go up or there's like a blackout and then it's, that's it and we're the backbone, we're 5% of the economy, but we power the next 95%. And we have to get this right. You know, let me talk about AI generally, and I'm going to take some questions from the audience. I mean, you're talking about it sort of as something you're powering, of course you must be using it yourself in some context. What are some of the tools you're using that are helping to both address these issues and grow the business. It's, it's a great, it's a great question, and I always say that you don't want your utilities to be the leaders in technology, really, why not? You don't, because when we lead and something goes wrong, bad things happen, so we, we want to be, I'll, I'll speak, I can, I'm speaking for Exelon, not EEI. When we, we like to be followers, we don't wanna be laggards, but we wanna be followers we cyber security is key for us because every piece of technology that you put on the system you make yourself more vulnerable. So we are always thinking about protecting the grid both physically and through cyber and we're also making sure our people are properly trained. I listened to the interview, I think it was Brad right before me and he was talking about people are the key to that change. And understanding utilities, it's not uncommon to walk into a utility and people have been working there 30, 40, 50 years, second generation. So that change management is second generation, aren't you? I am. You did your homework well. I mean, thank, thank you, uh, yeah, all of them. No, well, you know, talk about, talk about though with. AI, I would think that that's something we talk about as helping to basically alleviate and manage power. Are there any particular tools that you are comfortable with? I would say two things, and then I will tell you how I'm thinking about it in the future. The two things that we do is we like customer interaction. When you call in, if you ever need to call in, I want to be able to anticipate your needs. We're utilizing AI to do exactly that. 2, I'm probably angry because happy, but no, I'm just kidding, right? You don't call your utility when you're happy. You, you don't. And yeah, I, I've not had one of you call me and say thank you for my lights today, but that's all, that's all right. Noted. We're not working on that. Just wait till Monday. You'll be flooded. The, the second piece is making the regulatory and financial structure in which we operate more streamlined. We can take what takes months of just gathering data and, and we're utilizing that now. How do I want to use it? I talk about Exelon investing $38 billion into its system through 2028. We have a goal and we have stated it publicly that as a customer you should expect 24/7 reliability on blue sky days period. When the sun is shining, there's no rain, you should expect when you come home it's nothing more frustrating when your lights have been blinking and you're like what's up? Why? Shame on us. The second piece is after a storm, which is not a major storm, we should have 100% of our customers back online within 12 hours. And when we have a major storm, 100% of our customers back online within 48. That is significant. No one has ever said that's what they're going to do, but that's what you should expect from a world class utility, and I have to manage affordability. So how does AI come in? AI comes in and helping me anticipate where I'm having problems in my system so I can do proactive remediation where I can get in there and send you a notice and say you're gonna be down from 8. To 10 because we're working on this line that's not an outage that's me partnering with you as my customer and plus I mean it it is one of those industries where you have capped rates you don't have full, you know, um, I wanna take a question if there's any questions from the audience raise your hand, um. Yes, we have one over here and you can say thank you and, and you can please thank Calvin first and then tell us your name first. Thank you for the lights and everything else. Um, Ian Bramson Black and Beach, um, and uh my thank you for the partnership. You're welcome. Oh, OK, now this is a friendly question. My question's about cybersecurity. I run our industrial cybersecurity side. You had mentioned that earlier. You're talking about load. You're talking about. You have to go build stuff. Everything's at the forefront. We're putting in AI in here and not on the IT side. I'm talking about the operational side, keeping those lights on, keeping the AI centers running. What's the priority? What are you actually doing to build in fiber from the beginning? What's, what's in that? What, what are you doing there to make sure that that emerging threat is you're staying ahead of it? It's, it's all about governance protocols. Is early on we have we have we have a, a defensive depth strategy in all of our cyber security efforts that is embedded in any new technology that we're bringing into the system and we have protocols and governance procedures that have been set up way before we do any and. That just not goes into technology that goes into when we're building something we have to build it from the mindset that there's people out there that want to violate that system and they're trying to penetrate it each and every day so it is ingrained in everything we do, but what happens with that, it takes longer. Because now you have another sector that you have to incorporate into it and I'll be honest with you, the concern is not so much what we're doing my concern is by do my suppliers have that same ability to put the same protections on their system because once they get in they're in so we have to then go back to everyone else we're working with to ensure that they have the same protections and safeguards your comfort level right now with that given the number of. Stakeholders you have. It's not as high as I'd like it, um, because 1 to 10, I would say 6 or 7, yeah, because I would, the reason I say that is that I'm a big believer to give small businesses opportunities and small businesses do not have the capital and the infrastructure to put it in place. So therefore I have to put it's a belt and suspender type approach because I, I want you to grow with me, but at the same time I need to protect my customers and the system. You know, I, before, um, I give you some final thoughts, and if there's any other questions, like, forgive me, I just, uh, I'm, I don't see your hands. I'm gonna, um, I'm sure there's some out there, but. Um, we're talking about, uh, you know, you're mentioning almost really diversity in terms of bringing in some of these smaller suppliers, affordability, clean energy. How is the landscape around some of these other priorities, uh, shifting and especially clean energy because there's no question that, you know, we're on a trajectory where we have to care about that sort of thing. I would tell you that at Exelon it's still a priority because. When you I the jurisdictions that I outlined for you earlier, clean energy is still a priority nuclear and my, my view is that I'm going to partner with each and every one of my customers to achieve their goals and I've worked with administrations from Democratic to Republicans and the reality of it is is that. We are true to our mission and our values we're gonna be OK and that is how we outline it now the challenge that I do have is that because we're a 100% regulated business we need to get paid for all the work that we're doing and having those conversations up front matter. We're talking about 2026 predictions. Um, I know our time is winding down for final thoughts, but, uh, what's your prediction for prices next year? They're going to go up by how much double digit a lot. Look at exit to your right, so you're getting an out, but what, what do you, uh, any thoughts? Is it going to be something that is comparable to this year's increases? So let me, let me put it in perspective for you. We operate within the PJM marketplace. So again, the majority of us don't own generation. Last year, the governors in the PJM region put a price cap on. It saved customers $3 billion or the prices would have been higher. The governors have recommended extending that price cap for another cycle that would save anticipate another $10 billion. So that's why I can tell you with certainty that the prices are going to go up. So in the 39 seconds we have left, any final thoughts you want to convey to this audience of tech leaders, entrepreneurs, people forging ahead with the future of AI? Believe partnerships matter. We will get better because of our ability to work with you and if we do not get your voices into the conversation to make us better because utilities we know what we do but in partnership with you we can help technology make us more efficient and drive progress. I cannot think of a better place to end than there please join me in thanking Calvin Butler. Thank you very much. Thank you.

As OpenAI evolves from research lab to global enterprise, Chief Operating Officer Brad Lightcap sits at the center of its most pivotal transformation yet. He leads the company’s business partnerships—an area OpenAI sees as a key pillar of its future—along with its product strategy and path to profitability. Lightcap also oversees the OpenAI Startup Fund, which invests in early-stage companies advancing AI’s potential. He shares how OpenAI is scaling responsibly, sustaining innovation at breakneck speed, and balancing ambition with accountability in the race to commercialize intelligence.

The Business of Intelligence

Middle layers are shrinking, workloads are rising, and AI is becoming a new kind of teammate. This session will examine leadership changes when machines share the job of advising, coaching, and supporting teams. What skills will define the best leaders when employees no longer need to depend on a manager for every workflow? This session will examine and debate designing an organization where empowerment, not supervision, drives performance and value creation. 

When AI is Your Co-Worker and Your Manager

At a time when every investor wants a piece of the fast-growing AI pie, the landscape for emerging managers is thriving—and hotly competitive. Are their fresh ideas and focused approach here to stay? And how will they rub off on the rest of the VC industry?

Term Sheet Breakfast: Emerging Manager Vibe Check

Natasha Lyonne has built a career bending time, genre, and logic, from Russian Doll’s supernatural loops to her new frontier in artificial intelligence. The actor, filmmaker, and cofounder of Asteria Film Co. is using AI tools to create ambitious and visually rich films on independent budgets without compromising artistic integrity. Lyonne shares her perspective on how AI is reshaping Hollywood, what it takes to build the first copyright-clean filmmaking engine, and why the future of art depends on staying grounded, collaborative, and human.

The Human in the Loop

Smartphones, data centers, connected vehicles, wearables—is there any category in which SoftBank-owned Arm doesn't play? As the AI boom drives a need for energy-efficient compute in more places—and increases competition across the globe—Arm's chief executive outlines how the company's advantageous architecture and unique business model will drive it to new heights.

Within Arm's Reach

Renee, welcome to the Fortune Stage. Thank you. So excited to have you. So before we dive in, I want to give the room a sense of scale because I don't think everyone here is familiar with how pervasive arm is. You're in our phones, laptops, fridges, stoves, you name it. So give us a sense of how pervasive, um, arm is. Yeah, so what, what we do, uh, is, uh, a product called a CPU. Which is the brain of every modern electronic device. So I would say that each person probably maybe uses 50 to 100 armed CPUs on their person or in their home. I can assure you everyone has one today because we are. In literally every single mobile phone built we're in all the iPhones, we're in all the Android phones, but we're in Ford F-150s, we're in Teslas, we're in data centers, we're in washing machines, the brain that powers everything. So when you turn on your smart TV and all those apps come up and you're basically trying to pick whether it's YouTube or Amazon Prime. That whole operating system that's running on the TV, that's all running on an AM CPU. Uh, company's been around 35+ years. No, we're just talking. It's Cambridge University started in a barn in Cambridge with an ambition to be the global standard for CPUs, and that mission they accomplished it. I, I, I also want to double. It's not just devices you talked about data centers and you recently announced that in 2025 50% of all computes shipped to top hyperscalers will be ARM based. Tell us more about that because that that's kind of not expected. It's a little counterintuitive and also back to the history of ARM, um, we started in in a in a barn in Cambridge, and it was a joint venture between a chip company called VLSI Technology and Apple. And they were looking for the first chip for a first PDA known as the Newton if people remember that that device, which was a PDA way ahead of its time relative to running off batteries, had a display. There were a million things wrong with it, but what it needed was a low power CPU, something that could run off a battery, and that's how ARM was conceived, was a chip to run off batteries which fast forward served us incredibly well because back to the. To the mobile phones because the battery power is so key there and you're running off batteries, Energy consumption is key, power efficiency. So that is really the application space where we're quite strong. Fast forward to these modern data centers that are running 100 megawatts, 500 megawatts, gigawatts. Energy efficiency is everything. So ARM had gotten into the cloud data centers a number of years ago, general purpose cloud with Amazon and Microsoft and Google, but when Nvidia made the choice with their Grace Hopper chip and then Grace Blackwell, they needed a CPU to pair up with the accelerator and they chose ARM. And because Grace Blackwell is literally every single AI identity center compute. Being done, we're there too. So hence, hence our big number 50%. Incredible. Um, you've made a bold statement that ARM is the only compute platform delivering AI everywhere from milliwatts at the edge to gigawatts in the world's largest data center. What makes ARM uniquely positioned to play in this space? So let's think about the data center application we just talked about, right? These giant AI data centers that are using GPU acceleration, um, they all need a CPU as well because the CPU does all of the, the management of the data, prepares the data that goes in the accelerator. There's just a huge codependency there. You can't have a GPU without the CPU. That's fine when you've got. 50 megawatts to deal with 100 megawatts, but when you go to a very small application where wearable glasses, where you have maybe 2 watts of power, you can't put a GPU there. Simply put, you have to find a way to run that AI accelerator on the CPU platform. So that's where we play. So we not only have a solution up in the cloud, but we have a solution in the smallest devices. We've been talking about this all day, but I still would love a raise of hands if you believe we are in an AI bubble. Because perhaps you've changed your mind since the morning when Ali asked the question. OK, um, who thinks we're not in an AI bubble? OK, I think. And who, whose answer is like it depends, complicated. Yeah, I would say like 3rd, 3rd, 3rd. What do you think? Are we in an AI bubble? I think it depends depends on how you define bubble. If, if, if the definitions of AI bubble is that the stocks are overvalued and the PEs are running way ahead of themselves historically, you know, you know, maybe, uh, that's not something frankly I spend a lot of attention on. If you ask the question of are we an AI bubble relative to the technology being relevant and powerful, absolutely not, 100%. I feel that in my bones. And, and you might ask, well, why are, why are you so convicted about that? You simply just have to look at the, the areas that AI tends to help us with already. The the previous speaker was talking about vibe coding where you, you go in and, and you ask a question relative to write me a piece of software. Well, imagine a world where you go to AI and say, write me a business plan, right, or write me a way that I could come up with a novel concept to build a company that could solve this level of problem. Take life sciences. Uh, I've always thought that the killer app for AI is around health. If you look at one of the bottlenecks in drug discovery today, it's actually around trials, human trials. If you could replace all of that with AI and have AI instantly tell you from the time you're starting R&D that you know exactly what the trials are going to look like and you've accelerated drug discovery by 1015 years, that, by the way, all of what I described is going to happen. None of that is an if statement. It's a when statement. So when I think about AI and and the changes it can make for humanity and the planet and where we are relative to its capability, we're so early and the benefits will be so profound. So I, I'm absolutely not believing we're in any kind of bubble. Yeah, um, I'm gonna take audience questions in a moment, so please tee up your questions. You know, a lot of attention goes into the training of these AI models, but as demand for AI increases, the prediction is we're going to need more and more AI inference, and this is You know, the shift is going to benefit ARM. So tell us more about how you think about kind of the training inference. Yeah, and again, another example of why I don't think we're in an AI bubble. So just backing up for a moment, what is training? What is inference? Training is teaching a model to do something, uh, whether that is teaching you to recognize images, answer questions about history, solving an enterprise problem, etc. etc. Months and months and months of time spent training a model, training chat GPT. And OpenAI takes 3 months, 6 months, 8 months to train this giant model. That's training. Once the model is done, it's released to the world, and on our phones we can ask a question and the question comes back with an answer. Sometimes it thinks it takes some amount of time to give you the answer. That's inference. OK, so the giant use case as you think about it, of course, is inference. The more. Sophisticated the training model is, the more inference we're going to use. First things today, still researching companies spend more time training than they do on inference. So all this compute, these 100 megawatt data centers, they actually spend more time training than doing inference. And you say, well, why is that? Because there's just so much still that the models can get smart at. Inference is the use case that happens everywhere. It happens in the cloud. It happens in your car. It happens in your earbuds. And again, back to, you know, what is good for arm to your question. Those will all run in areas that generally will have an arm processor, so we think as inference explodes, and it will as more and more of these models get sophisticated, you're going to run these AI workloads everywhere. It will be ubiquitous. We will not think about AI running in the cloud or AI running in an edge device. It's just software. That's how it works. Do you have any questions? The audience, I'm not. Yes, there's a question at the back. Please share your name and organization. Sure, Julie Michelle Morris, I'm always here to ask about cybersecurity and the security of this, um, so. The single source, the supply chain, of course, makes me a bit nervous. Um, does it keep you up at night to know that the weight of the world is on your shoulders? Uh, we have a very unique situation in the semiconductor industry, and we were chatting about this, uh, uh, in the, in the green room prior, where, where chips were not so interesting for, for a number of years and then they got really interesting and, and one of the Events that made it so interesting was COVID, and when you lost your key fob, it took 52 weeks to get a new one. It was a bit of what is going on here. Well, that was just a function of the semi supply chain has many single points of failure, not just ARM. There's TSMC which is in a very obviously. Interesting part of the world geopolitically where it sits. TSMC is by far and away the provider of all these AI compute chips. Everything Nvidia builds is from there and AMD, Broadcom. That's a single point of failure. There is a very sophisticated device that has to go into these fabs that comes from one company on the planet for EUV. It's a company called ASML, Dutch company. They're the only people in the world to build EUV machines. There are a set of mirrors that go into that EUV machine made from a company in Germany called Zeiss. They're the only company in the world that makes those mirrors. So yes, I don't know if it keeps me up at night, but it is something the entire semiconductor supply chain has been learning to live with. But I will say, and you've seen this partially with the Trump administration, this acceleration to move manufacturing back to the United States, I think is a step in the right direction. There's a lot of criticism in terms of. How long it will take, and there's no question that it will take decades to get to where it was because it took decades for it to kind of get to the point that it is, but we have to start and I'm glad to see that it's starting because it's necessary. Well, when we were, yeah, there's a question here. When we were prepping for our call, um, I mentioned that I'm an investor in an early stage chip company that's very focused on AI inference. So perhaps you can give us a sense of like what does the competitive landscape look like. Well right now the leader is Nvidia by far and away, and they do almost everything relative to training and inference. Now when you have such a large competitor in such a large market that of course begets lots of competition, and there will be lots of competition from a combination of sources. You'll have hyper scalers doing their own chips, whether that's Google TPUs and that's gone on for a number of generations. Microsoft Meta AWS and then you'll have specialized companies trying to do chips around inference not only for the cloud but also for the smaller edge devices. So I would never try to second guess or underestimate Jensen and Nvidia. I worked there for many years. He's an amazing leader in an amazing company, but the opportunity is so large and so the scale is so broad, you will see other players, no question about it. And Iman and Kumar, so you know I do agree on the Nvidia side, uh, so building off of that, how do you see the hardware landscape evolve, right? So you know, grasshoppers, for instance, the bandwidth between CPU and GPU is increasing. So as these models get bigger, you know, agents, long context, it's all memory and bandwidth centric. So how do you see the. Ecosystem evolve and what would the, you know, CPU of the future look like? Yeah, no, it's, it's a wonderful question because. Ultimately when you're handling these large AI models and the models are if you enter chat GPT or enter a question into it or Grok and it says thinking and you're wondering well why is it thinking it's not the computer that's thinking, it's the computer that's waiting for data to come into the computer to figure out how to make the answer. That's memory bandwidth. So I think what's going to happen over time is you're going to see some innovations in memory technology. You're going to see innovations in packaging. You're gonna see innovations in interconnect. In other words, what's the material used to connect between the devices? Is it light? Is it a different kind of photonics? Is it a different area in terms of other interfaces? So I think back to your question on innovation, if you're, if you're investing in this space, it's a bit of the wild, wild west because people are looking for all kinds of different levels of innovation there because it all has to get solved. Selfishly, I think ARM is going to be in a wonderful place there because most of that traffic needs to run through or around our devices, and we're very involved in a number of activities in that space. Can we talk about physical AI, and that could be robotics, autonomous vehicles, drones, but also these small embedded AI first human-machine interfaces or wearables. And again, what does that mean for ARM? Well, again, because just about everything that you described is going to require something from a CPU standpoint to manage the activity, we will be involved. So then you get into physical AI, what's what's taking place there, and you see this a little bit with autonomous vehicles. In fact, I was reading an article that was in the New York Times this Sunday. It's, it was rather amazing. I was looking at Waymo and the number of accidents that Waymo had conventional accidents compared to human drivers. And it was about 1/10 to 1/20 just looking at and they were looking at different kinds of accidents, right, in terms of lane incursion or intersections. So there you have a data point that says, OK, autonomous physical AI is is starting to see benefits in in the cars and Waymo, wonderful company, they're all over San Francisco, is kind of autonomous 1.0. It's got all kinds of equipment on there. It's got LiDAR, it's got radar, it's looking all over the place relative to how to drive autonomously. The next generation using LLMs are they need less cameras, they need less devices, and why is that? Because they're actually using AI as opposed to looking at all the physical areas to make decisions on how to drive and how to how to move forward. What does that mean? It's going to really accelerate. Robotics, humanoids, things that can do factories, I think in the next 5 to 10 years, and again ARM will be involved, you're going to see large, large sections of factory work replaced with robots. And, and part of the reason for that is these. Physical AI robots can be reprogrammed to do different tasks. One of the issues you've had with factory robots in the past is if it was a pick and place machine for a factory, they're just optimized for one task for one task. The software was for one task, the hardware was for one task. Now if you can design a general purpose humanoid that the software is all AI and it learns by doing. It's going to completely replace a large set of factory workers. So I think in 5 to 10 years, and back to the geopolitical comment, I think it will level the playing field for countries because you will start to see physical AI be a great enabler for other economies. Quick question, because we're almost out of time. Renee, good, good comments. Um, AM is dominant. X86 is struggling. Risk 5, we'll see. Um, you talked about so many customers, right? Why then Is I'm going moving up the stack, um, going into building chips aren't you competing with customers, um, you know, what's the calculation there? Yeah, I don't think I said anything about that, um, 10, you know, it's a Silicon Valley rumor. Yeah, one of the things that we see is, is huge demand for our technology. And one of the other things we see is that the the speed required to develop products is just the demand on it is just getting faster and faster and faster. Part of that, and AI exacerbates it in the sense that. The models are moving far, far faster than the hardware can keep up. So when you sit where we do in the platform, in other words, the compute sort of starts with arm, it's not unnatural for customers to come to us and say, If the computer and models are all running on your platform anyway, instead of waiting for your IP, why don't you just build something for us that could be quicker. That is something we do listen to. So as I've said in our earnings calls very carefully, that is something we continue to explore largely because customers are pushing us for it. And why are they pushing us for it? They just want to get their end products out faster. Renee, last question, what are you most excited about for 2026? You know, I, I, I think that this continuance of, of AI and the acceleration of what we're doing, it to me, it's like, it's like being in Disneyland. It just opened, uh, we have hours and hours and hours to spend. Um, I was at a panel where. Someone said, give you a bold, give me a bold prediction for this time next year, and I said no one will be talking about an AI bubble a year from now. And then someone asked, because it popped? I said, I'm not gonna say big for that, but I, I just don't think we will. I think people have embraced a year from now that we're well on our way. I love that Disneyland just opened. Thank you for the conversation, Renee. Thank you.

Like many software companies, Intuit is rebooting its products for AI. Unlike many software companies, it’s helping Intuit steal sales from its competition. The Silicon Valley stalwart’s chief executive as well as its AI chief explain how the company’s AI ambitions have led to its fastest organic growth in a decade—and share where TurboTax, QuickBooks, Credit Karma, and Mailchimp might go next. 

CTRL-ALT-Compete

Thanks for being here. Thank you for having us. So. I'm from Philadelphia. All right. And we're kind of big on the Revolutionary War. That's kind of our, our, the name of our game, OK? And there's an old colonial American, uh, phrase around the revolution, join or die. We're all familiar with this. Yes. Um, Sussan, I feel like for you. And your company, it has been disrupt or die. Would you agree with that? And why? Uh, so we're a 40 year old company. Yes, we were born tech. Yes, in tech. We were born in the era of shipping CDs. That's when I joined the company. Oh nice. And so first mention the CDs on this, on this stage today. There are, there, there are lights in my eyes, so I can't see this crowd, but I'm sure some of you, maybe many of you, uh, were not even born in the era of shipping CDs. And so we, you know, we've been through 4 plus massive technology shifts and, uh, and many, uh, economic cycles, and, uh, I'll, I'll tell you the DNA of the company long before me has been about what you just said, live or die, uh, and so we've had to figure out how to disrupt ourselves, yeah, yeah, and, and here you are standing today. What's your market cap today? What's Intuit worth these days? I don't know what the stock market did today, but you know we're almost 190 billion. Not too shabby after 40 years, so cheers, um. Ashoka, I want to come to you. Thank you for joining us, um. So Intuit made a big bet on machine and deep learning, as we called it a decade ago. I said earlier on the stage, stays true for you, um. And my understanding is you really have to get the data ready to do something like that. Um, tell us a little bit about that journey because we're talking about getting the data ready now, but you guys were doing that a decade ago. That's right. uh, Andrew, it's a pleasure to be here. What I'd tell you is that a decade ago, as you said, we started to invest in 3 things. Number 1, building the right team, number 2, building the right data platform, and then number 3, building the right AI platform. These three things work very, very well together. The team has been trained on AI on how to use data, how to govern it, how to manage it. We've built the clean data pipelines that bring data to our AI systems at scale, but then we Really invested in connecting the data and the AI and then we built an operating system on top of this whole thing called GenoS which allows developers now to build massive applications at scale. To give you an idea, in the last year, fiscal year 25, we had 3500 independent projects and product experiences created on Genos. Yeah, that's remarkable. I would like to know if you changed your strategy over those, again, 10 years or so, you know, if, if you could have seen what was coming, hindsight 20/20, where you, when you set out to make these changes a decade ago. Did you pivot along the way, you know, just seeing where the market was going in terms of, in terms of AI and capabilities and so forth. So, um, when Susan took over as CEO and the executive team launched, one of the things that they defined was that Intuit is an AI-driven expert platform. That is the company's strategy. That strategy has not changed. What we did do though is start to evolve the technology in order to meet that strategy. And so we've made deep investments in clean data at scale. We made deep investments in building an operating system that could allow people to access that data and build AI experiences, and we were doing this before the Gen AI revolution. When the Gen AI revolution came, we started to make investments in creating that operating system. And that has really delivered measurable return on investment for our customers as well as for Intuit itself. Yeah, Sassan, um, Intuit's founders 40 years ago said that the company's biggest competitor was paper and, and the spreadsheet as we know it today. Is that still true? Uh, it is actually, and, uh, just maybe briefly, uh, for your audience, you know, we serve consumers, uh, businesses of all sizes, and, um, and we focus on things that matter to them, right, which is money, cash flow, uh, as I hear, I hear they're concerned about the money, yeah, yeah, right, but, uh, the, the reason I start there is, you know, you're talking about cash flow, you're talking about survival, taxes, bookkeeping, and accounting. And um and there's an incredible fear factor when it comes to how do I grow my business, how do I manage my business, uh, am I gonna have money to make payroll, and, uh, and you know, 75% plus consumers live paycheck to paycheck. So this is a real, you know, we talk a lot about AI, but there's like a real customer here uh that matters and uh what, what we need to do for them, which is how do we power their prosperity. And I say that just to say that the majority to your question, the majority of both consumers and businesses fall in a bucket that we call non-consumption, and what we mean by that is, um, for instance, businesses, uh, use 10 to 25 apps to manage their business. Most of the time their data and their insights are trapped in individual apps, and so they also have to have a tax accountant, a bookkeeper to help them advise them. Uh, and we, and they use Excel spreadsheets, Google Sheets, and so it's all non-consumption, meaning they don't use a platform to run their personal life or, or their business. That's, uh, and 40 years ago, yes, it was all paper. Uh, today it's just a different form of paper. It's all in apps, but it's all still trapped. The insights are still trapped, which is why. Those that we serve rely a lot on people and we believe people will still be a huge part of the equation along with AI, which, uh, I think it's counter to what many talk about depending on the industry they serve. So if, if they're really just looking for a better piece of paper, a better spreadsheet or or whatever it is, um, but really they're looking for a solution. To their problem, right? Let's be honest, um, and we talked about this earlier on, on stage with, uh, the CEOs of FreshWorks and, and Xero earlier today, right? What do, what do they really actually care about? Just as a solution to their problem. Um, why spend all this money on AI? I mean, it's whiz bang, it's, you know, really great. We can do all this extraordinary thing. If the bar is paper low, you know, why do all this? Is it just to gain customers from. You know, the, the, the, the, the clutches of, of paper and spreadsheets. I mean, you don't have to try that hard, right? That's right, or wrong. No, I, I love the question that, uh, that you're asking. First of all, um, I, I remind ourselves at the company all the time, customers don't care about AI, right? Everybody talks about AI, but the reality is a consumer is looking to increase their cash flow. A consumer is looking to power their prosperity to make ends meet. A business is trying to get more customers. They're trying to. Manage their customers, sell them more, more services, and We bet the entire company before this topic was popular more than 70 years ago on really 3 things. One, it's what Ashok talked about, which is data. Uh, we have to make sure that our data is clean, it's customers' data, not ours. It's clean, it's usable, it's digestible by the platform. The second is in order for us to grow. We declared 7 years ago we need to deliver done for you experiences. What does that mean? From lead to cash as a business, we'll help you manage your leads, we'll help you manage your cash flow, we'll do your taxes for you, your accounting for you. We'll give you advice on whether or not you should buy inventory and done for you for consumers from credit to wealth. Help you build your credit if you don't have credit to how do you manage your money and grow your wealth. And what we said was the only way to do that was data. AI and 7 years ago we said HI, human intelligence. Uh, if you look at our addressable market, the majority of it is, is influenced by people, people that are tax experts, accountants, bookkeepers that are advising you on what to do. Although the work is completely changing and we're automating it with AI, we believe AI plus HI is essential to reshape the category, which is really what the investments are that we've made. So really it's a long answer to your very short question. That's all right. Customers don't care about AI. What they care about is help me grow my business, uh, help me prosper, and we have found the only way to do that is to combine technology, automating everything for them. With human intelligence on our platform that can actually give you the human touch and the advice. And we believe that will be the case for decades to come, but the role of the AI, the human, will change. So quick poll for the audience, how many people have gone to the store for AI? I really need to buy some AI today. Zero hands. Andrew, to your point, if I could add to that, probably no one has gone to buy AI, but probably a lot of you want work done for yourselves, right? You want to have technology that do work that you would rather do, uh, have the machine do. What we've found is that with our agents on average customers are saving 12 hours per month with routine tasks. They would have had to perform. They're getting paid 5 days faster. They're 10% more likely to be paid in full. And as a person who's run small businesses in the past, I can tell you numbers like that are very meaningful. 12 more hours means 12 more hours that I can spend building my products, understanding my customers. And we have so many more statistics like this that show that the AI and the human intelligence combined together actually help people do better in their financial life. So this is really exciting. Can I just take 30 seconds and, uh, add to what Ashok just said, make, make it sort of real in terms of what a business experience is. So as I said. Customers don't care about AI. Uh, but and what we have launched is, uh, is a virtual team of AI agents that can hand off things seamlessly to a human that will do things like help our customers manage their leads, uh, help invoice and create estimates on their behalf. They're always in control, uh, to follow up with customers that have overdue invoices, write everything for them, tell them which customers, um, are overdue and which ones typically pay on time and which ones they could actually get paid earlier. So we do all of the work based on leveraging the customer data. And AI capabilities to do the work for the customers. So the stats you're hearing from Ashok, um, we have an AI, um, accounting agent that will do a lot of the categorization. That's the 12 hours a month they're saving. Our payments AI agent, customer doesn't know what it is. It's just it helps them with estimates, invoicing, getting paid, and that's the tangible stuff that where we believe we can. Change the world to do the things for customers that has a real economic impact around the world. Totally. I just want to call attention to the fact that Ashok has the letters A and I stitched into the right sleeve of his shirt in case anyone doubted him. Well, he has one stitch on his head, but we put makeup on. You can't see. All right, all right. I was gonna say I wanna see the tattoos later, gentlemen. Um, all right, so, uh, uh, I wanna ask one more question and then, and then we should go to the audience for them, um. Look, you guys have so many customers. How many? 100 million plus. So you have, uh, an, an unbelievable view into the economy of, of a certain segment. I know you serve all segments, but you especially, uh, serve small and medium businesses in terms of quantity. I would like to know. In your opinion, how they're doing in this economy. This is a very odd economy. Um, and you know, many of them fear AI or don't care about it or whatever. So what are you seeing across the board? Yeah, it's a great question. Uh, just brief context, out of the 100 million, about 10 million are businesses of all sizes, and, uh, and the businesses that we serve aren't concentrated in any particular industry, so we will serve plumbers, landscapers, construction, real estate, wealth management. I could go on and on, um, uh, the two big things that I would say is. Listen, the job market is still strong. Uh, credit scores, um, are lower than they've ever been, um, particularly with Gen Z. Uh, Credit balances are at the highest that they've been, uh, particularly with Gen Z. Their credit balances, um, credit card balances are up 36, 37%, but they still have jobs, uh, and that's what's really keeping things together. And so therefore everybody is being watchful about what they buy, what they don't buy, at what prices. Um, for businesses what we see across our base is the word I would use is stability, um, profits, cash flows are, are stable. Um, one element of it is they're on our platform, so it's biased data, um, digitization is helping them, but what I would say is industries like manufacturing, um, insurance are actually up. Uh, industries like lending, real estate are actually down in profit, so overall it's stable, but depending on country, states, and industry, very, very sector driven. Thank you for that. All right, let's go to the audience for questions, questions for the Intuit fellows. Yes, we've got one way in the back. Your name and who you're with, sir. Thank you. Hi. Hi, uh, Samir Gulati from Zen Business. Um, one of the hardest things about small businesses is actually getting started and the first couple of years, right? A lot of the things that he talked about don't even apply in many cases. Given your position in the industry, how do you think about using AI or the combination of AI and HI to enable that and make it easier? Yeah, it's a great question. Uh, you know, first of all, I'm inspired by what many are doing to help businesses get up and running much faster than they could before, but a couple of tangible examples that I would use, one is, uh, with our platform, uh, once you give us a few pieces of information that we ask for, like what, what are your goals, what are you trying to achieve, uh, and you know, for instance, uh, depending on where you are in your cycle, if you're 30 days in, if you're 60 days in. Uh, we will, in essence, ingest your data if it's website data, if it's data that you put together a business plan on an Excel spreadsheet, whatever it may be, if it's handwritten notes where you can take a picture of, we'll ingest, because first we need to understand what are you trying to do. We'll ingest that and then based on that make recommendations, uh, like if you're make it real, if it's a, if it's a plumber. Then we'll help a plumber with what are the basic things that they need, and then a plumber, they're on the road, they're trying to get business and so being able to estimate and create an invoice with progress payments to get paid right away is the capabilities we enable them with. So really what we can do now that frankly we couldn't. Do before is based on who you are, what you're trying to do, and the millions of customers that we have on our platform that are like you, we can help you with what to get set up with right away because you don't need all the capabilities that we have. Uh, what you need is what's specific to your situation. And then based on leveraging your data because we, we use machine learning, knowledge graphs, a lot of uh data, and of course G AI to help automate tasks, workflows and functions, but we can get you up and started right away and that also teaches us, by the way, what else you may need and that we um. are in the midst of making something generally available where you can do vibe coding and just say here's what I need done and we'll go do it for you. And so a lot of that just helps reduce the time to get started and to get your first customer and to get your first sort of cash flow in your bank account. And once we do that, then you're on a roll and that's really our focus in terms of the benefit for customers to get started. Thank you for your question. All right, gentlemen, I have two quick lightning round questions before we, before we wrap it up here. The first is, um, you've recently signed an open AI deal. Um, are you concerned about controlling your customer relationship given that? Uh, we're very excited about the deal. There are 3 principles that, that, um, are really important. Uh, one is the customer experience and the relationship. When the customer, when we're engaging the customer, they know you're, they're using the Intuit platform. Uh, the second is the customer's data does not leave our four walls. So it's our data, our models that are being trained. So the data and models and the way the platform, uh, behaves matters and just the economics have to matter and the economics are the same for us before and after the deal, so. And the thing that's very exciting with OpenAI is it's, it's the values and relationship we have with them, which is we both want to drive prosperity for the end customer, and I think it's gonna be a game changer in the long term. All right, and then my last one is I'm asking on behalf of everyone here and everyone watching on the live stream, when will AI be able to do my taxes, gentlemen? Well, I don't do AI and HI does your taxes today. We, we have, uh, you know, just to make it real, right, we have, um, millions of customers where AI does a lot of the drudgery for them and then hands it off to a human and the human finishes it for them. So we'll do your taxes today for you. Wonderful. There we go, gentlemen. Thank you very much. A round of applause for them.

Healthcare is moving toward a future where patients and clinicians both consult the same AI before they consult each other. How are we developing judgment, skepticism, and the skills needed to interpret machine reasoning? Healthcare professionals will debate what expertise means when anyone with symptoms can access diagnostic intelligence.

Your Patient Has the Same Algorithm You Do

In a world where every transaction carries both opportunity and risk, trust has become the defining factor in digital business. As AI embeds itself into global commerce, leaders face mounting pressures around privacy, resilience, and security. The winners will be those who can innovate quickly while protecting critical systems and safeguarding customer confidence, creating the foundation for adoption of next generation technologies. 

Trust as Currency: Winning in the Digital Age

AI’s compute appetite has triggered a global building boom. From Dubai to Dallas, developers and investors are racing to construct the next generation of data centers. These projects are about more than racks and cooling, they involve capital allocation, regulatory approval, and the geopolitics of energy and land. Where will the next hubs of intelligence rise, and who will control the infrastructure of the AI age?

The New Geography of Data Centers

The enterprise AI boom is moving from proof of concept to proof of profit. The chief of Google’s fast-growing cloud computing business explains how Fortune 500 companies, startups and public sector organizationsare building foundational AI platforms that blend intelligent agents, multi-cloud ecosystems, and next-generation infrastructure. He shares what’sworking, where adoption is accelerating, and how leaders are making the leap from experimentation to measurable business transformation.

Leading the Enterprise AI Revolution

Thomas, thank you for joining us. Thank you for having me. Um, we're gonna talk about agents, but not right away. Um, this is, uh, uh, a delight for me because, um, almost a decade ago in 2016, I was able to sit down just like this with your predecessor Diane Green and talk about Google joining the cloud computing business. And the whole vibe of that conversation was we are trying to transmit to the market, communication. To the market that we are serious about business now. This is Google, right? This is the consumer company, the company that we all know that it was a formidable thing and you know, as an Oracle, a longtime Oracle veteran that you know enterprise companies are built a bit different than consumer companies. So Today, Google Cloud is, by my understanding, you'll tell me, the fastest growing major cloud service, but it's off of a slightly lower base than your competitors, 13% of market share, uh, versus Microsoft's 20% and Amazon's 30%. Does that speak your truth? Yeah, OK, um. Why today are you winning more business compared to these other companies? Both because we have great products, people are finding them very easy to use, and they're getting real value from it. And we offer a very diversified product portfolio. Um, you know, our focus on cloud has always been, take amazing products, make them super easy for people to use, take care of the complexity that companies want us to deal with. Things like security, compliance, ah, but differentiate from everybody else across the strength of our portfolio. And so that's been why. And we've been, you know, we've had amazing growth rate. Um, Over the last 16 or 18 quarters, I think we've been the fastest growing. So if, if Amazon is the, you know, the the choice of, of the little guy, and that's how they get their volume, and Microsoft's supposed to be the choice of business, what is Google Cloud? We have a very diversified base. So we've done really well with startups, uh, particularly as startups shift to doing more AI stuff. We've taken a lot of share and. We're now the number one choice of AI startups by all independent reviews. Ah, we've done really well in traditional enterprises, banks, insurance companies, telecommunications companies, hospitals, ah, pharma companies. Um, and we've done well in not just The US, but in global markets, our revenue is highly diversified globally, which also people are surprised by how much strength we have in overseas markets, and that's reflective of the we've got a very, very focused plan on how we build and take our solutions to market. We've We also got great partnerships with many independent software vendors. We've got huge practices. I think there's 700,000 consultants who implement Google Cloud globally, which is a huge number. And so we've all that's come because of the demand we see from customers. Sounds like an enterprise business. One of the other things that struck me about what you're doing, and this is very on trend last couple of years, is allowing the use of any model, right? So when you choose to, when you, when you go with Google Cloud, you don't just have to use Gemini, you really have a, a wealth of options. That is that right? Yes. OK, so how in that circumstance, which is great, how in that circumstance do you competitively differentiate, right? Because you're not the only one who, who, who makes that, that offer. Yeah, so we, we offer. Our differentiation when you look at AI is to do AI really well, you need a stack of technology. And a stack is energy, chips or systems, infrastructure, models, tools and applications. And we're the only ones who offer all of it. We also offer a choice at any layer. So for example, we have our own amazing silicon called TPU as some people call it. Ah, but we also work very closely with Nvidia because depending on the kind of model you're running, you want that choice. The same thing, if you look at how people use our platform, ah there are customers who use a combination of models. Uh, Intuit, for example, the customer, they use some of our models, Gemini and our diffusion models, Imagine and VO, but they also use other models from other people for tasks that they want to use. I would not say that because you offer a stack, it's closed. I think many other companies don't have parts of the stack, so they depend on others for silica and other people for models. We have our own, but it doesn't mean it's closed, and so you get the best of both worlds. You said closed, reminding me of the open versus closed debate, which is usually in the context of the models, but you're, you're speaking more broadly than that. Is that debate dead? We've had it here before on this stage. Is the open versus closed dead? You're a tech guy. In enterprise, you have to have, for example, if you look at our models, they can access information from any application, database, collaboration tool, because in enterprise, you have to have it work with everybody. And so we've always had an open philosophy, uh, back in 2019 when we said, Companies will want to have cloud technology from multiple providers. The incumbents said no, they would not want it. Today, 95% of large companies use cloud from many, many different organizations, and so that gives them choice. And we've always felt for enterprises, they want choice, and choice means open. So I'm gonna ask you a, a big question and we're gonna have to unpack it a little bit, so stick with me, um. When you got started, we were calling this cloud computing. AI was not quite on the tip of our tongue the way it is today, um. We were calling it machine learning and deep learning I think back then, um, solely um. How have you had to rethink this service given these massive waves that have happened since for, and this is the unpacking part, stick with me, Thomas. The AILLM arms race, the AI demand boom from your own customers who suddenly all want to be AI companies or use it, and then the actual capabilities of AI itself and how that's changed what you do. So I'll, I'll keep you on track, but Let's start with the LLM race. How has that changed what you offer and how you do what you do? So we, we've been working on AI a long time before LLMs came along. That's right. And we've always designed for the long term. So if you look foundationally, when you look at what we were trying to solve for cloud computing, it was take technology and make it accessible to everybody by making it super easy for people to use it. And with AI we've always had the same philosophy. So, at the foundational level, we said people want amazing systems to train models, but we at the same time, we also realized you can't just train forever, you actually need to do inferencing well in order to do, to make money as a company. So we built systems for both training and inference. We also knew that The the most problematic thing that was going to happen was going to be energy, because energy and data centers were going to become a bottleneck alongside chips. So we designed our machines to be super efficient in delivering the maximum number of flops per unit of energy. So that's one example. When we built flops per unit of energy, yes, it sounds like a soccer metric, but I, I digress. So, the second thing we felt was, when you look at LLMs, you can't just hand over an LLM, you need a tool set around it, uh, in order to make the LLM really work for organizations. So that toolset needed to secure the LLM, it needed to access information, it needed to help you tune it. Uh, it helped needs to remember what you're asking it and handle long-running things. So we built an entire tool chain around it, and we gave it to developers. And to give you a sense, when we launched Gemini 2.5, uh, we went from 0 to 1 trillion tokens in like 7 hours. Because it's so easy now to get it and scale it. Uh, so that's the 2nd thing. Third thing, we always felt that along with those who want to roll their own as developers, there's always going to be parts of the organization that don't have the capability to roll their own. For example, a legal department that wants to use our models to understand uh contracts and which contracts have indemnification and some warranties. Uh, marketing people who want to create advertising and really tailor the advertising to the segments, and they want to create the advertising content using a model. Uh, for all of them, we built a portfolio of applications. Uh, we took Gemini, which is our app. And made it available for companies by integrating it with all the enterprise systems they have and we built some domain specific ones that have gotten great, great traction. For example, we built one to help people find products online, uh, because today when you search, I have never found any product online. Never. No, I'm kidding. I'm kidding because you're not used to it yet. I'm kidding. The internet is an addictive thing. Well, look, if I may distill, um. A lot of iterative building for to to to get what your your business ready for these successive waves. I wanna, I wanna go back to something you said earlier about power so. We are all well aware of the finite amount of power that we can generate on this planet. I'll go so far as to just let leave country, you know, international boundaries behind. Um, are you concerned about that? Yes, you've been building for efficiency, but the adoption curves have been. Massive and fast. I mean, we've never seen anything pick up like this before. Are you concerned about your ability to power your business, your ability to keep, you know, pace with the, the, the, the amount of demand that you and your competitors are seeing? Take us into your, uh, uh, your brain as far as the, uh, power side of thing is concerned. We are working on a variety of new technologies that would you like to share any here? Well, I. Tell you that there are 3 parts to what we're working on. One is to be as diversified as possible in the kinds of energy you can use to power computation for AI. Uh, it's, most people say, can you use any form of energy to power AI. That's actually not true, because if you're running a cluster for training and you bring it up and you start running a training job. Uh, the spike that you have with that computation draws so much energy that you can't handle that from, uh, some forms of energy production. So one is diversifying those. Second, being super efficient in how we reuse energy within the data center itself, so that you're not just bringing energy once in the data center, but the thermodynamic exchange can be super efficiently done and used. And third, some new fundamental technologies to actually create energy in new forms, and we're working on all three of them. Since you often ask, where are we applying AI, we're using applying AI, the control systems that monitor the thermodynamics in our data centers are all governed by our AI platform. The dog fooding, as we like to call it, right? Um, let's talk about the, the TPUs, the custom silicon that that Google is working on. Again, another trend, all the big tech providers, the hyper scales are very into the custom silicon thing. What role does that play for your business forward? You're already using it. You have been for a while. Um, are you trying to, you know, eliminate all of the chips from other vendors or is it in certain circumstances, where does the, the TPU roadmap go vis a vis your business? We've, I mean, we've worked on TPUs since 2014. That's right. So a long time before AI was fashionable, we were working on, on the silicate. And we kept investing in it because we fundamentally believed that you can change the design of an accelerator, which is what a TPU system is in fundamental ways. Uh, we offered that to our customers through our cloud. Ah, many companies are using it. Many of the top labs use it, many of the unicorns use it. And we are now broadening it to support new forms of computation. So for example, capital markets firms are switching from numerical computing to inference, and they're using it. There's people in high-performance computing, there's people in molecular simulation, energy simulation. So there's a variety of different domains. And at the same time we worked very closely with a number of companies on adopting and making our technology work amazingly well on GPUs, most importantly Nvidia. We've optimized Gemini and our models on it, and so that's been a long term partnership as well. Um, I wanna ask you one more question, but I, I, I do want to come to the audience as well. So, uh, just give me one more and then I'll come to you, um. I was going to ask you about that very, that very company in fact. So, uh, Jensen, Jensen Wong, the CEO of Nvidia, um. If he were sitting here right now, would he say that he was Excited and thrilled about the TPU roadmap that Google has, or is he concerned that it's going to harm his business? You should ask him. Oh, wow. No, but truly, I mean the, the big push for custom silicon. The big push for custom silicon threatens the business of, of these, of, of the, of the other chip makers. Obviously you need a chip maker to make the TPUs, a contract chip maker. Jensen is a great partner of ours. We've done work for a very long time. You know, I'll, I'll give you an example of the kind of work we've done. For a long time, if you wanted a, uh, a model that was uh an LLM that had a closed weights and didn't share the weights. But you had information in your data center that you wanted to hook up to it, and you wanted to bring the model into your data center, you could not do it. He asked, he gave us the idea and we worked for 3 or 4 years now to really optimize Gemini. And you know, earlier this year, we announced that you can take Gemini and run it on an Nvidia GP cluster anywhere. And we have to do it while protecting our model weights and our intellectual property, and that's an example of a joint project. We work with them on compilers, workload management, there's many, many things. Because I think there's there's way too much in the press on. It's a zero-sum game, when in fact, for those of us who have been working on AI infrastructure, there's many different kinds of chips and systems that are optimized for many different kinds of models. And as the market grows, we're creating opportunity for everybody. Yeah, yeah, understood. All right, let's go to the audience for questions. Forgive me, it is rather bright. Yes, right up here in the front, wait for a mic, and then your name and who you're with. Thank you. Hello. There we go. Hi, I'm Ash from Adeline. Um, Google works with a lot of customers through solutions engineers and getting them to implement LLMs within their organizations, customers, enterprises. And some of these customers, I'm sure they go through the journey of trying to build something but don't come to the other side, meaning it doesn't work for some reason or last minute it gets pulled back. What are the reasons from your perspective that enterprises today still struggle to push something out, either as an internal tool or to their customers, and what will we need to do as a market to kind of get them over the hump? Thank you for that briefly, Thomas. Typically, it's one of 3 things. They didn't design the architecture right. Second, they didn't clean up the data. Third, they didn't test the model and think about things like security, and if the model gets compromised, ah, how to handle it. Uh, and lastly, we've seen often, they didn't, because they didn't think of the problem the right way, they didn't think about how to measure the return on investment on it. There you go. Check yourself before you wreck yourself. One more, please. Your name and who you're with. Hi, Julie Morris, um, thought leadership for cybersecurity, etc. Um, thank you so much for this. I'm curious, so Google Cloud acquired or Google acquired Wiz, um, biggest acquisition, um, but there's still this big gap around model security and Uh, I'm curious, is Google Cloud going to try to step into any of that to do more? We, I think we are working on two different dimensions on, uh, with AI. One is how to protect against threats that are created using AI models, and second, how to protect AI models themselves. Uh, because a lot of company-specific data is now inside these models, and people are always worried about somebody hijacking them and taking access to the data. We have a lot of deep science work going on at Google on both dimensions, and you'll see us present more and more of it as we go forward. There you go. Thank you very much, Thomas. Thank you so much for coming back to Brainstorm. A round of applause for him. Thank you. Thank you so very much.

Want generative AI that actually delivers results? It's harder than it looks. Two leading AI startups that specialize in corporate readiness explore the models and techniques tech leaders must consider to deploy effective AI systems. Open source or closed? Fine-tune your LLM, or retrieval-augmented generation (RAG)? And just how good must your organization's data be? We dig in.

From Rag to Riches

Great guys, so we're going to talk a little bit in this session about you know how enterprises can really get AI to work at scale and deployments, and a lot of that has to do the title of the session is from Rag to Riches and rag here is retrieval augmented Generation, which some of you. Mayo is a technique for helping AI get the data it needs and the context it needs to give good answers, but we know, you may not be familiar with my 22 panelists. So Dawa contextual AI, just 30 seconds. What is it? What, what do you guys do as a company? Yeah, so you mentioned rag. Reg is really about making Gen AI work on your data, right? So I think one of the The interesting thing about where we are right now is we have these amazing language models and amazing G AI, but the question is how do you make that work on your data, and the problem to solve there is the context problem. How do you give it the right context? So that's what we're focused on enterprise startup, uh, solving hard problems with context engineering, essentially. Got it. And, and Arvind, similarly, Glen, for those who don't know Glen, what's Glean do? Yeah, well, I mean, think of GLEA like GPD but inside your company. It's a place where your employees go and ask any questions or give it work, and GLEN will use all of the world's knowledge as well as all of the internal company's context to, to actually answer those questions or complete those tasks. And so in many ways, you know, I think, you know, the underlying technology, the concept that we use is similar, which is understand your enterprise, understand your people, your data, your knowledge. And bring that as useful context, ah, combine that with the power of language models to actually solve, you know, real tasks and answer real questions. And you guys have seen like tremendous adoption, right? You've just hit 200 million in, in annual revenue run rate, is that right? Yeah, that's right. So we just a couple of months back, you know, we, you know, announced that milestone, I guess we did today, but the, uh, but that's exciting, you know, business is growing for us, uh, rapidly. Great, um. So Dawa, you sometimes talk about sort of RAG 2.0, that there's this kind of new way of doing retrieval augmented generation. What is that really and how is that? How are people doing rag today that might be different than how people thought about doing it, uh, maybe right after chat GPT came out in, in 2023? Yeah, or even before that, right? So RAg is older than, than chat GPT. So, um, so my claim to fame originates is, is I'm an AI researcher, right? So I was the project lead on the original RAg project at Facebook AI Research. And at the time, the way we did things, vector databases didn't exist. The model we gave the data to was not called a language model. It was just a sort of generative model, but it wasn't a language model. So if you look at where we are now compared to then, it's just a very different idea even. So, so the core framework is still the same. You're still doing retrieval to give context to the generative models so that it can deliver value. Uh, but the way you do that is very different. Um, and that's why when we started contextual, we were really obsessed with this context problem. I still think that's the most exciting, most interesting problem right now. So we have these amazing language models, even if you freeze all technology, and you just say, OK, this is, you know, whatever GPT 4 5.2 or Gemini 3, we just keep that. There's still a decade of innovation that we can do just from focusing on the context. Um, and so the, the way you really want to do that is not just a vector database, right? There's a lot more going on. We're, we're doing agentic rag these days. We're using retrieval as a tool, and there are many more tools that you can use. So that's why we, we were looking for a new. We called it RAG 2.0. I don't know if that was the right name really, but I think the core observation was that what you're building there is a system. So it's not a model, right? The language model is maybe 10, 20%, and so it's the same for something like Glean, right? That's an extremely complicated system. And the language model is just a very small part of that, and I think that's why we need to have something like a rag, right? But then if you want to do that the right way, then you have to start from the systems assumption, and that's what, that's what we've been doing. And if you do that the right way, then you get much, much higher accuracy. So that's really what we're seeing kind of across the board is. Enterprise AI adoption is all about high value use cases and if you really want to get there, then you need high accuracy and that's what we do much better than anybody else. Interesting. And, and Arvin, for getting this to work well, Enterprise search is key. How, how do you do that successfully? And when you, uh, you know, there's a lot of data out there that a lot of enterprise, uh, uses of, of AI are not really scaling. They're not getting full deployment. What do you think the pitfalls are there? Why, why, where are companies going wrong? Yeah, I mean, see, first of all, enterprise environments are very, very complex. Um, there is lots of information, lots of data across hundreds of thousands of different systems. Data is permission, it's governed, like not everybody in the company can access our data. So when you sort of start naively, and you say that like, look, I'm gonna actually make AI work in my enterprise. I'm gonna connect AI with all of my systems, that's when you start to sort of run into. all these challenges. Why AI is so hard. Like, first of all, just connecting all of the enterprise information, like bringing that context into AI is hard because that means you need to build integrations to all these different enterprise systems that you have, understand the data models, understand the APIs, you know, to connect with that data. Then the second challenge you run into is that, well, oh, you know, I, I launched AI and I connected it with my Um, with my documents, but you didn't actually take care of security. And you have now AI having access to really sensitive information inside your company, and everybody in the company can then go and use it. So, so like when you, like, you know, like this, this, you know, we're all in the very early stages of making AI work. There was like this tremendous pressure from the executives, you know, uh, team, you know, within the company. Um, we had to do something, make AI work, and we didn't really, they properly like sort of prepare. For it and how to, how to make it happen correctly for us like at Glean like you know we've been in business for now almost 7 years, so we're just like, you know, like you know you're one of one of the original sort of enterprise AI focused companies. The, the first application that we deliver to our customers is, is search, which is, you know, we go in and connect Glean with all of our enterprise systems and we sort of understand more robustly, first of all, you know, what is all this. Information that you have inside the company, what information is up to date, uh, is high quality, what information has become obsolete at this point, because, you know, when you think about context again, you have to make sure that you're using the right context and making AI work on it as opposed to making AI work on scale. How much of that sort of data cleaning can be automated and how much of that is still kind of a manual process? Someone has to look at a human has to look at it, say, yeah, this is out of date. We're not actually going to put that in the database that we allow the AI system to access. Yeah, I mean, I think like when all these projects failed, like, you know, like if somebody needs to take the blame, so who took the blame, you know, the, uh, the startups, you know, and the product companies will tell you that, hey, your data sucks, you know, it is bad and it's your fault, right? So go and fix it. But the reality is that humans cannot fix, you know, the data. You cannot suddenly change our, our habits. You certainly can't like get people to. You know, actively maintain, manage information, keep it up to date, remove the steal information out. That's not how we work. Like, you know, it's, you know, we're going to work the same way as we work today. And so the, the right solution is for AI itself to figure it out. AI has human capabilities, and if you have the right AI, it will actually you know what information is good and what is not, and it will actually make careful use of, like, you know, you know, the information the right way and again, like, you know, the right security. Do, do you agree with that? Is that you can do that. Yeah, and, I, I think one of the most sort of misunderstood sayings I often hear from people is garbage in, garbage out, which is total BS. I think like what, what you want is garbage in information out, and that's why you want to use AI in the first place, right? So I think that is a very common misconception. So maybe we're not quite there yet, but In the long run, what you want to do is just apply AI to your data and all data in an enterprise almost by necessity is going to be garbage. So the question is how do you get the useful information out of that? Absolutely. I'm going to try to get one question from the audience in a second. So think of a question and I'll hopefully get. At least 11 in the audience, um, there's a lot of talk out Arvin, uh, uh, about sort of context engineering as like the next frontier of how to get Enterprise A out of work. What's your view on that and, and how are you, how is Glean working with things like MCP, you know, and scripts and all of these new ideas about giving more context to models? Yeah, I mean, I, I think that that's exactly like the new frontier of innovation is like for any given task, like how do you. build the right context and context is dynamic. Like it's not that like, hey, I have this task, and you suddenly just like you go into all these systems and pull some right pieces of information and you're done. It's actually iterative process where, you know, models and agents are going to be solving complex tasks and Every step of that execution, you have to sort of like figure out like what's the best way to bring the best information to these agents so that they can do their work properly. Um, so, so that's, that's, that's absolutely the key. Like, you know, what we do at GLEAN is, we are, I mean, we think of ourselves as we're trying to build, you know, your enterprise brain. Um, the idea behind GLEAN is you connect Glean with all of your different systems inside the company. So now we understand. You know who your different people are, um, what are the different business processes, projects, teams, uh, product lines that you offer, so sort of building a deeper understanding of your enterprise and also work like, you know, what are the different business processes and how do individuals go and actually execute those business processes today without AI, like, you know, what are humans doing? So sort of if you observe, if you actually connect with all the systems and you observe all the human activity, you're able to sort of then bring, you know, that collective human intelligence. of enterprise in one place and you do that and this is basically the repository of context and now from here, like when you actually ask AI to do some, you know, some piece of work you've got to sort of now, you know, from that repository pick, you know, the right context, you know, and that right context really is like, you know, well how did humans in the past did similar things, you know, in this company so that AI can sort of use that as the as the baseline to start and then do do its work and that's how it sort of really improve accuracy and. And and build these agents that are uh that are capable of actually extending, you know, that human intelligence and, and, and, and automate work. Alright, I wanna get, I wanna get one question for the audience. Who's got a question? And if no one does, I've got, oh, right over here is a guy, uh, with a question. Please state your name and where you're from. Rick Chavie with VX, uh, enterprise scaling platform. One question we, we are getting with our customers is around. If they have a lot of M&A and they've got multiple ERPs, e-commerce, CRMs, whatever, is coming up with that consistent framework unified to be able to apply the agents or any other technology for that matter to give that sort of single view of, of revenue execution or or whatever application you're looking at. So I'm, I'm wondering how. With your models, you know, how you sort through all that confusion, or in some cases systems may not even be really automated. You have to, yeah, and it's, do you need a single source of truth, or can you work across multiple systems? You should work across multiple systems again, so I, I think in a, in the complexity of an enterprise you are not going to have everything in one place. Right, so that's why we have things like MCP, some of the information you might want to index so that you can do faster retrieval over it, but, but in the end I think you just have to accept that you are dealing with a sort of messy reality of an enterprise and now you need to have an AI system that can deal with that. So part of that is understanding what you need to look for and so you know, doing the right tool calls, hooking up into the right ERP systems. A lot of the complexity that that that. that you solve with something like a glean or also with our platform is really about hooking up all this data. Now you have this, so you have this kind of data layer then you have The intelligence layer and then the missing part is the context layer. Like how do you get all of the right little bits of context from that in an iterative way so that intelligence can be useful. Arvin, I, I want to play something you said. You talked about having the AI learn the existing context and the existing workflows to sort of replicate those, which is probably a good place to start. But there's this tension there because a lot of people want what they really want to do with AI is reinvent workflows and change workflows. Can you do both at the same time? How, how does that work? Yeah, in fact, actually what I was saying was, see, there is that human intelligence, the human way of doing things today in your enterprise, and that's the, that's the starting point. So when you, when you, when you ask AI to actually solve a new problem, like either, you know, you let it loose, you know, with no sort of, uh, baggage, you know, to deal with, so to speak, like, you know, like let AI actually go and figure out. From, from scratch, like how to solve that task, or you say that, well, look, you know, like, you know, there's, there's some value in you knowing, like, you know, how we've been doing this work in the past. And you actually bring that context in and then, but you don't like, you know, you don't ask AI to sort of replicate that exactly what the, you know, what, how you were doing that business process before that actually just serves as the inspiration, the starting point, and AI can actually make its own decisions, you know, on top of that. So, that's how it works. So like, you know. Most of the use cases like as people are building these agents today, what we're seeing is that AI converges faster and is able to add value and you know, sort of like, you know, conforming to like you know how the business really wants to work if you bring in that, you know, context. Interesting, um, I wanna do a couple quick rapid fire rounds because we're, we're running a little low on time. Um, if what do you think is the most overhyped sort of, uh, AI technology at the at the moment? uh, what, what do you have to say? Agents, agents, Arvin. I, I, I, I, I would say that any technology, you know, any cool demo, like, you know, it seems to make a complex business process somehow like, you know, magical and automated right away because none of them is gonna work. Interesting, but I think on agents, so I think agents. Is most overhyped but also probably most underhyped. It depends on the time scale you're looking at. Interesting. I was, well, that's my next question is, you know, in, in the next year, what do you think's gonna come along? What, what, what, what's on the roadmap that you think's gonna make the biggest difference, to kind of enterprise implementation, Ivan, what do you think? Well, I mean, I, I think the models like are delivering amazing capabilities every year. Ah, this year was a big one for, for us in terms of the, you know, the, the Like the level of agents, the complexity of tasks we can handle through the thinking models. So I actually don't know what's coming next year, but again, like, you know, to, to what you said before, I think we're very happy with the underlying model technology. And, and now it's time for us to sort of really build robust integrations into enterprise systems and building a deeper understanding of. Um, your enterprise processes and automate them, but for me, like what, what, what's most exciting is that next year AI is going to become proactive. Today, like most of the AI use inside the enterprise is reactive in the sense that I, as an individual, have to go to AI, make it, ask it to do something for me, and it does it, and very few people do that. People don't have time to actually make it happen. So next year I think we're gonna see the, see this flip and AI. OK. Do, what do you, what do you see coming? Yeah, I, I think test time compute to your point is, is really, I think a lot of people are underestimating what this means. So this is really going to change everything. We are just getting started, um, and so when you hook up all your data and then you have these smart reasoning models, uh, that can do so much more, we are going to unlock all of these very complex use cases that really seemed like science. Fiction and that's already happening everywhere now. I mean, like even with things like coding, right, but how good we are at that now with AI. I, I don't think I would have expected that even a year ago. So, uh, it's going to come very quickly. I, I think next year is really going to be the year of, uh, of ROIs and through complex. Well, we're, we're out of time. We'll have to invite you guys back next year to see if your predictions are right. Thank you very much for, uh, for joining us. Thank you. Thank you.

AI breakthroughs are moving out of academic labs and into companies that are training and  deploying the most advanced systems in history. That shift is reshaping who drives scientific  discovery and how quickly ideas make an impact in the real world. Three leaders at the forefront  of this transition explore how their work in health care, climate science, and autonomous  mobility could redefine the way we protect people and the planet. As the balance of power tilts  toward industry, who decides which innovations scale first, and whose needs they serve? 

The New AI Power Shift

"Move fast and break things" has long been a Silicon Valley mantra, but the rise of AI has us rethinking every component of the modern corporation. What's mission-critical and what's not? Can you be too lean and mean? And: In an Intelligent Age, is killing organizational complexity theultimate competitive differentiator? Two Valley veterans discuss.

Death to Complexity

With Alexa+ and a new generation of Echo devices, Amazon hopes to turnthe smart home intosomething more intelligent and intuitive—andperhaps teachits old virtual assistant a few new tricks. Panos Panay discusses how the company's Alexarevamp aims to transform a growing constellation of connected devices into “ambient AI.”

Home Is Where the Smart Is

Machine learning may promise efficiency, but most of us are still drowning in back-to-back meetings and digital overload. Who has time to take a mental pause? Calm CEO David Ko offers a rare moment to explore how thoughtful, human-centric approaches to AI can support digital mental-wellness tools. He shares how the leading mental health platform is exploring the integration of intelligent tech to help executives sleep better, think clearer, and lead with more intention, while navigating anxiety and stress along the way. The session closes with an AI-guided meditation to help you do more than just power through.

Calm in the Chaos

Everyone, thank you for having me here today. I um, I'm excited to be here at Fortune's Brainstorm AI. Um, over the last 3 years, I've had the privilege of leading comms, uh, really transformation. And that transformation process has gone from us being a consumer, a global consumer app, really focused on sleep and your mental health. To being now focused more on and transforming more into a full end to end mental healthcare company. And this work for me is extremely personal. I had my first panic attack when I was 14 years old. I had sweaty palms, tightness in the chest, uh, shortness of breath, and I didn't really understand quite what was happening at that moment. So for today, and for me today, it's much more than a job, it's a mission. And this mission is to help as many people as humanly possible. To access mental health services on a global basis whenever they need it. Now, uh, I'd love to make this a little bit more interactive. And the way I'd love to make it a little bit more interactive is that we all know that it's really hard for leaders to lead in this new age of AI when we feel that our batteries are depleted. Now, we know that mental health is probably the most approachable that it's ever been. I love that we can have these conversations on a global basis. But here's the thing, why can't we have these conversations at work? It's not like the TV show Severance, where you go to work and your mental health just checks itself at the door. So I wrote a whole book about this and I tried to take a heavy topic and make it a lot more approachable at work. And that was to equate your mental health like a phone battery. It's much easier to say to someone, how's your battery, than it is to say, how's your mental health, right? So, quick show of hands. How many of you today feel that your battery is at 75% or above, just raise your hand. OK, cool. How many of you today feel that your battery is at 50%? Raise your hand. All right. And 25% or below. OK, a couple of you. Thank you for that. Let me talk a little bit about why the, the numbers that you all gave today kind of follow what we've seen. We did a survey of over 250 C-suite executives. And in that survey, when we asked them, how are you doing, it followed most of this room. Most people said that they were doing good. That's actually kind of how I've been when I've talked to my employees in the past in hallway conversations, when I've asked them, how are you doing? And the reality is most of us want to just hear that you're doing good. We just want to hear that you say I'm doing good. If you say I'm not doing so good, you're like, oh wait, I have to go to a meeting. I didn't leave time and space for that, so I'm a lot more intentional today than I was back then. And here's the thing, in our survey, when we asked people how are you really doing, that's when we start to get real responses, and we found that over 48% of those individuals were feeling overwhelmed. We also saw that 25% of them were feeling anxiety or depression. And that only 1 in 4 executives said that their batteries were fully recharged at the moment. So let's talk about the why. And what's happening in this great paradox that we see. How many of you in this room have trouble sleeping, sleeping at night? OK, that's a pretty big audience. What we found was that when we asked our executives, how many of you have trouble sleeping at night, and it's not the quantity of sleep, it's the quality of sleep. So many of them at 41% said, hey, we struggle at night having good quality sleep. A little hack for you all if you can do this. If you can actually not let the phone be the last thing you see at night, that's a great first step, before you go to bed. And a bonus, if you don't let the phone be the first thing you see in the morning, that can help your quality of sleep immensely. Just give it a try. The second thing we found is if you have if you can't sleep at night, the next thing that we started to see is that 34% are feeling drained. And then 40% are unable to be mentally present. So think about this, what's happening. You can't sleep at night, you're feeling drained, and you're not mentally present in the office. That leads to a startling fact, that 1 out of 2 who took our survey actually said that they are thinking of stepping down. These aren't just data points, everyone. This is kind of the new leadership reality that we are living in today. And here's the thing. Most leaders are like in this room. are operating at about 20%. Think about what that means. If you were an athlete today and you wanted to go out and do a sport, would you go out and try to do your best at 20%? If you were a coach, could you really lead at 20% that team? I think for most of us intuitively, we would understand that answer is no. So right now, we are making some of the biggest decisions, whether it's in the office, whether it's around large topics, whether it's around AI, the future of work. At 20%. So if you really think about it, These conversations are so important. And here's what else that's happening today. We're having conversations in AI that are super broad. We're having conversations about how AI can make us more efficient. We're having conversations around AI will it replace humans. We're having conversations like in today about what is the future of AI and work. But from where I sit, At the intersection of technology and mental health, I believe we are asking the wrong question. The question is not whether AI will replace humans. I think the thing that we have to ask ourselves as a leadership group is, how do we support humans so we can lead in the age of AI. And I'll say this a different way as well. How can we support you, so you can lead in the age of AI? And let me show you some stats around this. Our surveys have showed that 79% of the leadership teams believe that the human brain is the original data center. That makes sense. But we've also seen that 13% say they fear that AI will replace human workers, which is lower than the general public. And most leaders feel that AI is actually very complementary. This is where our conversations around mental healthcare and mental support are so important from a foundational level going forward. And this is some of the work that we're doing at Calm and Calm Health. And let me tell you what we're doing. We're seeing when you engage with calm around your mental health that you're 2 times more likely to engage in our mental health services. And outpatient care. We find that over 77% complete a mental health screening. We find that 39% will actually enroll in a clinical program. And 33% of folks say they actually felt better after one session in terms of their overall well-being. And what I'm sharing with you aren't just stats. They're really a new foundation for us moving forward. Companies invest in their mental health really see a 3 times higher, larger engagement. They see less burnout and a higher ROI. They see 4 to 1. 84% of folks believe that mental health directly impacts the bottom line, and most leaders believe it's 71% that healthcare services like mental health will be embedded in their programs going forward. So in a world that's currently being transformed by AI, organizations are realizing that our greatest assets aren't just the technology, it's the people behind them. Now here's some quick ways that I like to recharge because they don't have to be these big grandiose gestures. They can be simple things. You look out the window, you do walk, you do water. The surgeon General once said, just check in on someone, Vivek Murthy, the former surgeon General that is. So you just focus on one small task at a time. And then lastly, it's not just about talking about how to recharge. Let's talk about how you actually do it. I actually have a really short AI guided meditation that was done by ChibzoKrake that I'd love for you to hear. If you can stay seated, open and close your eyes, just stay relaxed, and let's take a quick listen. Hey there, my name's Chibzo Karrike. Let's do a quick breathing exercise to help you reset. Maybe your mind has been going nonstop, weighing decisions, navigating pressure, and holding everything together day by day. Outwardly you might seem calm, but inside the stress has been mounting. So let's take a moment to pause. You can close your eyes or just soften your gaze if that feels better. Try bringing your awareness to your breath as it moves through your body. Breathing in, the body expands. Breathing out, the body softens. Breathing in Breathing out. Now take a moment to notice what it feels like when your mind gets a small moment of space. Reflect on the qualities you want to bring forward. Steadiness. Clarity. Presence. Even in an AI driven world, it's your human capacity. Your judgment Creativity And empathy. That lead the way. OK, go ahead and open your eyes if they were closed. Wishing you a clear mind and steady energy for the rest of your day. I hope you, uh, I hope you all feel recharged. I'd be remiss to tell you we have free calm at the front with QR codes. Enjoy your day, everyone. Thank you very much.

Brainstorm AI San Francisco 2025

Rivian's Next Charge

Latest video

With production ramping up and a new wave of models in the pipeline, Rivian is entering its most critical chapter yet. CEO RJ Scaringe talks about how AI, autonomy, and bold engineering are powering the next evolution of the company and the EV industry itself.

Videos (37 videos)

Rivian's Next Charge

Finance in Flux

The Future Makers

Reinventing the Enterprise

Brains for Bots

Bots on the Block: Your Lunch is Learning

Leading San Francisco: A Conversation with Mayor Daniel Lurie

The Next Phase of Governance

The Business of Creativity: Where Innovation Meets Impact

AI at Velocity: Securing the Agentic Enterprise

The $100 Billion Data Question

Defense Disrupted

A New Kind of One-Stop Shop

The Future of AI Infrastructure

Scotty, We Need More Power

The Business of Intelligence

When AI is Your Co-Worker and Your Manager

Term Sheet Breakfast: Emerging Manager Vibe Check

The Human in the Loop

Within Arm's Reach

CTRL-ALT-Compete

Your Patient Has the Same Algorithm You Do

Trust as Currency: Winning in the Digital Age

The New Geography of Data Centers

Leading the Enterprise AI Revolution

From Rag to Riches

The New AI Power Shift

Death to Complexity

Home Is Where the Smart Is

Calm in the Chaos