Hello and welcome to Eye on AI! In this edition…Wall Street’s leading AI skeptic is at Goldman Sachs; Hollywood celebs urge Gavin Newsom to sign SB-1047; A highly anticipated AI startup comes out of stealth; and (surprise!) Russia is using AI for election influence efforts.
The president of the United Arab Emirates met with President Joe Biden yesterday to hammer out a deal to cooperate on AI issues. The geopolitical backdrop of the meeting was certainly complex—the Gulf nation wants easier access to U.S. technology, but the U.S. has restricted Gulf states from freely importing U.S.-made AI chips over concerns the technology could leak to China.
Observed from another angle, however, the meeting was quite simple: Money makes the world go round. The U.S. wants—and needs—Middle East oil money in order to remain dominant in AI. And Gulf states like the UAE, which want to become AI powerhouses themselves in a bid to diversify their economies, are more than happy to shower U.S. AI companies with billions.
Sam Altman wants to drive down AI costs
Of course, wealth may be good, but who doesn’t love a bargain? Sam Altman certainly does. The OpenAI CEO has been talking for months about reducing the costs of the computing power and infrastructure that is fueling the AI boom. Yesterday, he published a blog post that, while ostensibly focused on the untapped potential of AI that will drive “massive prosperity,” clearly challenges the world to build enough data centers and manufacture enough AI chips to make developing and operating AI cheaper.
“If we want to put AI into the hands of as many people as possible, we need to drive down the cost of compute and make it abundant (which requires lots of energy and chips),” Altman wrote. “If we don’t build enough infrastructure, AI will be a very limited resource that wars get fought over and that becomes mostly a tool for rich people.”
Middle East wealth poised to feed the need for AI power
Of course, cheap often doesn’t come cheap. Altman knows this, given the news in February that he was seeking trillions of dollars to reshape the entire landscape of AI chips and data centers. In the meantime, the Gulf States, including from sovereign wealth funds in the UAE, Saudi Arabia, Kuwait, and Qatar are seen as one important way to feed—and fund—Altman’s need for AI computing power, and that of many other hot AI startups like Anthropic and Mistral. OpenAI is rumored to be in talks with MGX, the $100 billion UAE-backed investment fund, for a multibillion-dollar funding round that is said to value OpenAI at $150 billion.
CNBC reported this week that funding for AI companies by Middle Eastern sovereign funds has increased fivefold over the past year, though Saudi Arabia’s human rights record has led some companies to turn down its cash. Anthropic reportedly ruled out taking money from the Saudis in its last funding round, citing national security issues.
There are also AI partnership deals flying around the Gulf: Ahead of the UAE’s meeting with President Biden, UAE AI company G42 announced it had teamed up with AI chip juggernaut Nvidia to work on climate tech. In April, Microsoft also announced a $1 billion investment in G42, and last week the two companies said they would open two “responsible AI” centers in Abu Dhabi. Last week Saudi Arabia’s Aramco partnered with AI startup Groq to build the world’s largest data center for high-speed AI processing.
The environmental cost of AI is alarmingly high
Money may make the world go round, but more down-to-earth concerns like air, water, and food are essential to humanity’s survival. When it comes to the environmental impact of AI, the real price of the massive investment in AI chips and data centers—aimed at reducing the cost of computing power required to harness AI—is alarmingly high.
The Washington Post reported recently that OpenAI’s ChatGPT must chug just over one bottle of water to write a 100-word email. One email per week for a year would use up 27 liters of water and that means if one in 10 U.S. residents asked ChatGPT to write an email weekly, it would require more than 435 million liters of water.
Cheap AI will cost a fortune
Is it all worth it? When seen through Sam Altman’s rose-colored glasses, the answer is an unequivocal yes. In typical overly dramatic, prophetic tones, Altman wrote in his blog post that “I believe the future is going to be so bright that no one can do it justice by trying to write about it now.” (Sorry, Sam, I’m still going to try.)
Costly downsides like AI’s environmental impact warranted zero mention in Altman’s dispatch. They also seemed to be largely overlooked in the geopolitical discussions about AI’s future, including the recent meeting between President Biden and UAE President Mohammed bin Zayed Al Nahyan.
Only time will tell how these issues will play out. But I get the feeling that Sam Altman—and other AI and global leaders—already know that the quest for cheap AI will cost a fortune, in more ways than one.
Before we get to the news: I highly recommend checking out Jeremy Kahn’s latest Fortune magazine story that dropped this morning. It focuses on how Intel’s years of missteps have left it fighting for survival in the Nvidia-dominated AI era: Who could have imagined that Chandler, Ariz., outside Phoenix, where Intel is investing nearly $30 billion to build two state-of-the-art semiconductor plants, would be where the company’s fate might be decided?
Sharon Goldman
sharon.goldman@fortune.com
AI IN THE NEWS
Goldman Sachs’ stock research chief has become Wall Street’s leading AI skeptic. In an interview with the New York Times, Jim Cavello, head of stock research at Goldman Sachs, discussed a widely read research paper published three months ago in which he challenged whether businesses would see a sufficient return on what by some estimates could be $1 trillion in A.I. spending in the coming years. His take: Generative makes so many mistakes that it may not reliably solve complex problems. Skepticism from Cavello and others has led to what some have called a turning point for AI stocks in recent months. In the piece, Cavello challenged the notion that AI costs would decline and also criticized AI's capabilities: “Overbuilding things the world doesn’t have use for, or is not ready for, typically ends badly,” he said.
Hollywood celebrities sign open letter urging California’s governor to sign SB-1047. More than 125 Hollywood actors, comedians, writers, directors, and producers signed an open letter urging California Gov. Gavin Newsom to sign the hotly debated SB 1047, which would regulate developers of AI technology. “We fully believe in the dazzling potential of AI to be used for good,” the letter says. “But we must also be realistic about the risks. Now is our moment, Governor, for California to set that practical and positive example for all of the United States, as well as the rest of the world.” The letter comes a week after comments from Newsom that he is concerned about the possible “chilling effect” on the industry.
A highly anticipated AI startup comes out of stealth. TechCrunch reported that an AI startup backed by a who’s who of angel investors emerged from stealth today with $10 million in seed funding. Letta, which says its AI models remember users and conversations, is backed by heavyweights including Google’s Jeff Dean, Hugging Face’s Clem Delangue, Runway’s Cristóbal Valenzuela, and Anyscale’s Robert Nishihara. The startup was created in University of California at Berkeley’s Sky Computing lab, which was launched in 2022 to “accelerate the transition to cloud computing.”
Russia is using AI for election influence efforts, say U.S. officials. It may not be surprising, but U.S. officials have confirmed that Russia is embracing AI in its efforts to influence the U.S. elections in November. As per NPR, an official from the Office of the Director of National Intelligence told reporters that AI is making it easier for Russia, as well as Iran, to quickly tailor and target polarizing content meant to sway American voters. “The [intelligence community] considers AI a malign influence accelerant, not yet a revolutionary influence tool,” the official said. “In other words, information operations are the threat, and AI is an enabler.”
FORTUNE ON AI
LinkedIn, Facebook, and Instagram are hoovering up your data to train their AI. Here’s how to stop it —by Jenn Brice
Marc Benioff blasts Microsoft’s Copilot AI tool as the next ‘Clippy’ —by Sasha Rogelberg
Sam Altman and Apple’s legendary designer Jony Ive are collaborating on a secret AI project —by Chris Morris
Cloudflare is arming content creators with free weapons in the battle against AI bot crawlers —by Sharon Goldman
OpenAI’s newest creation is raising shock, alarm, and horror among staffers: a new logo —by Kali Hays
AI CALENDAR
Sept. 25-26: Meta Connect in Menlo Park, Calif.
Oct. 22-23: TedAI, San Francisco
Oct. 28-30: Voice & AI, Arlington, Va.
Nov. 19-22: Microsoft Ignite, Chicago
Dec. 2-6: AWS re:Invent, Las Vegas
Dec. 8-12: Neural Information Processing Systems (Neurips) 2024 in Vancouver, British Columbia
Dec. 9-10: Fortune Brainstorm AI San Francisco (register here)
EYE ON AI RESEARCH
New survey by Harvard professor says American workers are adopting generative AI at a faster rate than earlier technologies. About 40% of working-age Americans already use generative AI tools, with about 24% using it at least weekly, according to a new survey from David Deming, a professor of political economy at Harvard’s Kennedy School, and researchers from the Federal Reserve Bank of St. Louis and Vanderbilt University. The survey, which was conducted in August and claims to be the first of its kind on a national scale, pointed out that it took nearly a decade for the personal computer to reach a similar level of adoption by comparison. “One important thing to remember is that fast adoption was only possible because nearly everyone has access to a computer and to the internet,” Deming said in a statement. “These technologies build on each other.”
BRAIN FOOD
Judi Dench will speak to you now. Meta plans to announce it has secured deals with actors including Judi Dench, Kristen Bell, and John Cena to lend their voices to the Meta AI chatbot, according to Reuters. The new audio feature will let users select a voice for the digital assistant, with both celebrity and generic voice options. The reporting comes in advance of Meta’s annual Connect conference, which focuses on the latest developments in AI, mixed reality, and wearables at the parent of Facebook and Instagram.