If we want to truly “become an economy based on producing real goods and services,” as Greer told Congress, ending the trade war for US manufacturers needs to be the highest priority.

The outlook for the manufacturing sector — that part of the economy that tariffs were supposed to help — is bleak. Manufacturers continue to shed jobs (down 88,000 year-over-year) while productivity collapsed in the fourth quarter of last year, the opposite of what we would expect if tariffs were boosting productivity, as Greer alleges.

And it’s factory owners themselves who are telling us this story. Manufacturing sentiment, a measure of how manufacturers feel about the growth prospects of their sector, remained negative for most of 2025. In fact, The Economist reported that manufacturers overwhelmingly reported negative sentiments when tariffs were mentioned, with no respondents reporting positive views. There’s little to suggest the manufacturing sector is clamoring for more tariffs.

Notably, the administration seems to be focusing on production instead of prices, likely because the data show that tariffs are raising costs for consumers and businesses. Research from the Federal Reserve Bank of New York and other academics suggests that importers have borne about 90% of the tariffs, with about a quarter of that being passed onto final retail prices shoppers pay. This will continue to worsen if the tariffs stay in place.

When businesses have to pay more for their manufacturing inputs, and consumers have to pay higher prices for imported goods or purchase higher-priced domestic alternatives, the economy slows. This is felt beyond just factories. More spending on tariffed goods means less spending elsewhere, like on services, causing those other sectors to shrink. Tax Foundation modeling shows that tariffs in the long run will actually reduce GDP and lead to a smaller economy overall than if the tariffs had never been imposed.

Though it will take years to fully comprehend the magnitude of the effects the first year of Trump’s trade war had on the US economy, the topline estimates for 2025 do not indicate that tariffs boosted overall growth in the short run. Real GDP for 2025 grew by 2.1% from the annual level, lower than the 2.8% growth in 2024 from the annual level. This is roughly in line with what the CBO projected prior to the new tariff regime.

To be fair, tariffs are only a portion of the economy, and other forces — like the government shutdown — took a toll on growth. But if we look at real final sales to domestic purchasers, a measure that is not confounded by reductions in government spending, we can see real final sales declining steadily in 2025. We can’t say definitively that this decline is due to the president’s tariff policies, but if the tariffs were strengthening the economy, we would expect the opposite trend.

Greer cites the president’s trade and investment deals with various countries as evidence that the tariffs have allowed the US to extract significant concessions from its trading partners. But the deals that have been announced are at best frameworks for future negotiations, including pledges to expand market access or invest in the US, with no enforcement mechanism. Not one of the deals has been ratified by any of the legislatures in any of the negotiating countries, including the US.

Even just looking at the data for 2025, we see little evidence of an investment spike from the tariffs. Foreign direct investment was lower in 2025 than in the previous four years. And most of that FDI was reinvested earnings, rather than new investment.

If the tariffs remain in place, over time, we would expect to see some shifts in supply chains to the US. But that would not be a costless adjustment. It would involve raising prices significantly for businesses and consumers, while also damaging US credibility with our trading partners.

Greer told Congress we are in “a moment of drastic, overdue change.” He’s right: the economy is changing, with the AI boom and other emerging technologies creating new markets for countries to dominate. Yet the president is pursuing a trade agenda that treats the country like it’s 1926, not 2026. America should be pursuing policies that ease the burden on US manufacturers to become the dominant force in this race, not increase it.

The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.

This story was originally featured on Fortune.com

Concerns over the near-$39 trillion national debt burden the U.S. Treasury has accumulated over time are rising up the political agenda. The debt, accrued under both Republican and Democratic administrations, is now costing more than $1 trillion annually in interest payments alone. Both the president and his Treasury Secretary, Scott Bessent, have indicated they’re mindful of the debt issue, suggesting tariffs and visas as revenue-generation schemes to help balance deficits.

Debt hawks aren’t convinced by their plans: Many are backing a policy target to reduce deficits to 3% of GDP, around half of where they are now.

One influence many look to is Phill Swagel, the director of the Congressional Budget Office (CBO), whose projections are relied upon to see how the debt question may play out. Swagel is heartily optimistic on the matter, saying he believes a debt crisis will be avoided entirely because he has faith Congress and policymakers will act before a reckoning comes to pass.

Dimon seems less convinced. He told Tangen at the NBIM’s Investment Conference in Oslo yesterday: “The level of things that are adding to the risk column are high like geopolitics, oil, government deficits. And they may go away, but they may not, and we don’t know what confluence of events causes the problem.

“So if you look at all economic history, it’s a different confluence of events, different tectonic plates hitting each other. And they may affect 2026, they may not, but they need to be resolved. And if they’re not resolved properly, they will cause real additional problems down the road.”

A question of inflation

Depending on who you ask, there’s a range of outcomes for how the debt issue will shake out. Some, like Dimon, expect a market reckoning when bond investors begin to raise their rates, as they view the U.S. Treasury as a riskier borrower as the debt pile grows bigger. Others, like Bridgewater founder Ray Dalio, see the issue first cropping up in public spending, with programs being squeezed out by the interest payments the government is handcuffed into paying.

Another is the notion of “financial repression,” whereby the government allows (or is forced to allow) inflation to take hold in the economy in order to reduce the real value of the debt. Financial repression also comes with the idea that the government could encourage banks into holding its debt at lower interest than the rates set by the market—a further inflationary pressure.

Dimon said his inflation expectations are broader than most of the markets (this will be of no surprise to those familiar with Dimon’s ethos—he’s known to run wide-ranging risk simulations to ensure America’s largest bank would survive).

He explained: “My view is that there are a lot of inflationary things out there, including the Iran war, the remilitarization of the world, the infrastructure needs of the world and our deficits. I ask all my economists … I don’t know how the world running deficits like this isn’t inflationary. And you just may not have seen that yet. That die may have been cast, it just hasn’t happened yet.

“So when I look at scenarios I’m looking for early indicators, but it is possible that inflation ticks up and that will catch a lot of people off guard.”

This story was originally featured on Fortune.com

• Markets: Oil up (again), stocks all over the place.
• Exclusive: CEOs got millions after boards ‘neutralized’ the impact of tariffs on exec comp.
• Exclusive: Nvidia’s Jensen Huang heralds new era of productivity powered by AI.
• Exclusive: AWS chief Matt Garman sees “massive change” ahead.
• Musk in OpenAI case: AI “could kill us all.”
• Trump signals blockade of Iran will continue for the long term.
• Facial recognition cameras at Disney.

Quick note: Subscribe to the forthcoming Fortune Gulf Brief. Every Tuesday, this new newsletter will deliver clear-eyed, authoritative intelligence on the deals, decisions, policies, and power shifts shaping one of the world’s most consequential regions, written for the people who need to act on it. Sign up here.

This story was originally featured on Fortune.com

The San Francisco-based security and compliance company has crossed $300 million in annual recurring revenue, Fortune exclusively learned. This milestone represents a tripling of ARR in two years. Vanta’s customer growth rate has also accelerated to roughly 60% year-over-year—a number that has gone up in each of the past four quarters, the company said. Sources with knowledge of Vanta’s balance sheet say that its net revenue retention (NRR) has similarly increased every quarter for the last 2 years and continues to be over 100 percent.  

Vanta now serves more than 16,000 customers, including Snowflake, Atlassian, Duolingo, Ramp, Cursor, and Harvey.

The company’s success stands firmly against the backdrop of its last public valuation in July 2025: $4.15 billion. At the time, Wellington Management led a $150 million Series D alongside Sequoia, Goldman Sachs Alternatives, J.P. Morgan, Craft Ventures, Y Combinator, Atlassian Ventures, and CrowdStrike Ventures. Vanta has raised more than $500 million since CEO Christina Cacioppo and engineer Erik Goldman founded the company out of Y Combinator in 2018. Cacioppo—who taught herself to code from books before writing Vanta’s first prototype—now oversees roughly 1,000 employees.

What’s pulling the curve up is a problem that didn’t exist at scale 24 months ago. Vanta’s own data, drawn from its third-party risk management product and released in a recent report, found that 70% of companies now have shadow AI—tools employees adopted without security review. The company also reported that LLMs are 52% more likely to be flagged as critical risk than traditional SaaS. In a single year, the average company sees employees reinstall an AI tool 1,000 times after security has revoked it. The most-reinstalled offenders, per Vanta: Claude, ChatGPT, and Cursor.

“There’s this push-pull going on at an actually really broad scale,” Cacioppo told Fortune, describing executives demanding AI transformation while security teams scramble to assess what’s already on the network. The fix, she argues, is continuous monitoring.

“AI is exciting, but also scary and risky,” Cacioppo said. “It’s that combination of new, quickly growing AI hyperscalers with more risk and more scrutiny that is letting Vanta’s growth rate actually increase year over year.”

Now, the $65.2 billion governance, risk, and compliance field is watching a category leader extend its lead. As for the inevitable IPO question, Cacioppo demurred: “The goal is the long-term sustainable company versus the day of confetti.”

See you tomorrow,

Lily Mae Lazarus
X: @LilyMaeLazarus
Email: lily.lazarus@fortune.com
Submit a deal for the Term Sheet newsletter here.

Joey Abrams curated the deals section of today’s newsletter. Subscribe here.

This story was originally featured on Fortune.com

The times today are very different. The shift from encyclopedias to artificial intelligence isn’t just a technology upgrade, it’s a fundamental change in how we interact with knowledge. It’s reshaping how we work, how organizations operate, and how opportunity gets distributed. That’s why AI is becoming such a defining force in this next chapter of work and life.

The evolution of knowledge tools

If you look back, there’s a clear progression in how we access knowledge. In the encyclopedia era, information was static and curated. It was reliable, but finding and interpreting it took time.

The search engine era changed speed and access. You could type a question and get thousands of results instantly but the responsibility still sat with the individual to evaluate, synthesize, and decide the relevance.

Then came the platform and data era. Software organized information into dashboards and workflows, giving people real-time visibility. Decisions became more informed, but humans still had to interpret the data and turn it into action.

Now we’re entering the AI era. These systems not only retrieve information but also help make sense of it, including analyzing, summarizing and, increasingly, doing. Instead of spending hours searching, you can ask a question and get a structured answer, a recommendation, or even a draft of the work.

Across these stages, the role of the individual has evolved from being a researcher to a navigator to a data-driven decision-maker, and now to a director of intelligent systems. It’s been a subtle shift, but an extraordinary one in terms of how work gets done.

What this means for work

Healthcare is a good example of this shift, especially given how high the stakes are.

Not long ago, physicians relied on training, textbooks, and journals. As information and knowledge expanded, staying current required a significant investment of their time. While the internet made information more accessible, it also made it more overwhelming. Doctors still had to sift through large volumes of information to find what mattered.

Electronic health records centralized data and added digital support, though this digitization also introduced an administrative burden that often pulled physicians away from patients.

AI has the potential to reverse all of that. Clinical copilots can summarize histories, identify patterns, suggest diagnoses, and handle documentation in the background. When this works well, it gives physicians back what they value most: their time and attention to focus on judgment, trust, and patient care.

More broadly, this shift is about more than just productivity gains. When people spend less time on tasks that drain them and more time on parts of their work that require human judgment and connection, work becomes more meaningful. It brings them closer to why they chose their profession in the first place.

This isn’t just a nice idea, it shows up in the data. The HP Work Relationship Index finds that, when people have access to the right tools and technology, they’re five times more likely to have a healthy relationship with work, and 69% say they’re excited about how technology will improve their work experience.

The rise of the individual enterprise

One of the most interesting aspects of AI is how it expands what an individual can do. Historically, organizations existed to bring together different types of expertise. Complex work required teams of people with varied backgrounds, such as analysts, researchers, and operators, because no one person had all the capabilities and expertise.

AI lowers that barrier. People can now access tools that support research, analysis, writing, planning, and execution. In many ways, it allows all of us to operate more like small enterprises, supported by intelligent systems.

You can see this in practice. A physician can reduce administrative time and spend more time with patients. A small-business owner can run marketing and analytics without a large team. An entrepreneur can launch and scale with far less infrastructure.

The productivity gains are enormous, but just as important is the sense of agency this creates. People have more control over how they work, what they focus on, and how they bring ideas to life. And with that comes a greater sense of ownership and fulfillment in the work itself.

Implications for society

When access to knowledge changes, opportunity tends to follow.

The printing press expanded literacy.
The internet expanded information.
AI has the potential to expand expertise.

For the first time, people almost anywhere can access capabilities once limited to large organizations. That’s powerful—but the benefits won’t be evenly distributed.

AI will likely widen the gap between those who learn to use these tools well and those who don’t. Skills such as judgment, creativity, communication, and ethical reasoning will become even more important as information is easier to generate.

Today, that gap is already emerging. While nearly half of business leaders report using AI tools daily, only about a quarter of knowledge workers do. Access and adoption will be just as important as the technology itself.

Education will need to adapt. The focus can’t just be on memorization. It needs to shift toward problem-solving, critical thinking, and learning how to work alongside intelligent systems. It becomes less about what you know and more about how you apply it.

The human element

With all the attention on AI, it’s easy to lose sight of something simple: technology works best when it supports people, not replaces them.

AI can process information faster and at greater scale than humans can. It can surface insights we might miss. But it doesn’t build trust, show empathy, or navigate complex human situations with judgment and care. Those remain uniquely human strengths and they matter even more in this environment.

What this shift really changes is where we spend our time. Instead of focusing on finding and organizing information, we can focus more on understanding it, applying it, and making better decisions.

If we navigate this thoughtfully, AI can make people more capable, organizations more effective, and society more innovative. It can also make work more fulfilling, giving people greater agency, more room to focus on what matters, and a stronger connection to the impact they have. The tools are changing, but the goal isn’t. It’s still about expanding what people can do and what we can achieve together.

The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.

This story was originally featured on Fortune.com

For the 2023-2024 school year, the government provided 4.8 billion lunches to the nearly 29.4 million students belonging to the National School Lunch Program, at a cost of $17.7 billion, according to the U.S. Department of Agriculture data. Part of this sum takes the form of cash reimbursements to schools serving free or reduced-cost food to students, with free lunch costing roughly $4.70 per student per meal.

Many schools, however, say the assistance they receive to feed students the subsidized meals are not enough. A recent survey of more than 1,170 school nutrition directors from the trade group the School Nutrition Association (SNA) found this year, 69.6% reported insufficient reimbursement rates to cover the cost of school lunches, an increase from 67.4% the previous year. More than half of the directors said there is “serious concern” about the financial sustainability of their school nutrition programs over the next three years, up from 46% from the 2024-2025 school year.

Changes to school lunches

These school nutrition directors were surveyed in October 2025, and since then, additional factors may threaten the robustness of school lunch programs. Though reimbursements per meal increase each year alongside rising food costs, President Donald Trump’s One Big Beautiful Bill Act slashed funding for the Supplemental Nutrition Assistance Program, ending automatic free meal eligibility of children. Fewer children qualifying for SNAP lowers a school’s identified student percentage of those requiring assistance, meaning fewer reimbursements may be offered to schools providing free or reduced-cost meals.

The Department of Health and Human Services, under Secretary Robert F. Kennedy Jr., has made a concerted effort to address the quality of school lunches, pushing schools to eliminate ultra-processed foods in favor of whole foods, fruits and vegetables, and more protein. The push is part of Kennedy’s Make America Healthy Again movement and retooled dietary guidelines released at the beginning of the year. Schools must abide by these guidelines, intended to address children’s health, in order to continue receiving federal funding.

While nutrition experts have lauded the push toward scratch-made meals and more whole food options under the updated dietary guidelines, economists worry the push will further strain schools already concerned with the future of their school lunch programs

“The issue here is the operational reality of getting there with the current level of funding,” David Ortega, professor of food economics and policy at Michigan State University, told Fortune. “Not having enough staff, culinary training that comes with trying to do a lot of that more whole-food scratch cooking, the need for equipment and infrastructure—these are really operational issues that have to be addressed from a funding perspective.”

Last month at a Congressional briefing, SNA President Stephanie Dillard called for increased funding to help schools accommodate the new guidelines.

The “$4.70 must cover the food and the supplies, our labor and our equipment, deliveries and utilities, and the list goes on,” she said. “Our St. Patrick’s Day menu featured a scratch-prepared  Shepherd’s Pie using fresh, locally sourced beef.”

“I wish we could offer this option regularly, but local beef is priced about $3 more per pound.” 

Ortega said that without increased resources, the cost of serving affordable or free meals to students will increase, forcing schools to make difficult decisions.

“You may have loss of coverage for some children and schools, and schools may be taking on more debt for some of the unpaid meals,” he said.

HHS and USDA did not respond to Fortune’s request for comment.

The rising costs of feeding children

The cost of feeding schoolchildren has been on the rise for years, in tandem with rising labor costs and increased food prices. Since early 2020, grocery prices have increased nearly 30%—far outpacing the broader inflation rate—as a result of the pandemic, geopolitical conflicts like the war in Ukraine creating supply chain bottlenecks for commodities, as well as a tightening labor market in agriculture and food production driving up wages and therefore food costs. The preexisting challenges were reflected in the survey results showing rising concern for the financial sustainability of school meal programs.

“These programs were already stretched thin before the guidelines changed,” Ortega said.

But the new dietary guidelines present additional challenges because of the resources needed to execute them in schools already strapped for resources. The updated standards advocate for more red meat and whole-fat dairy, but beef prices have been on the rise amid dwindling cattle herd sizes. According to USDA data, more than half of schools report food service staffing shortages, which may require schools to invest in more expensive equipment to improve kitchen efficiency or risk not feeding students in a timely manner.

If MAHA’s goal is to truly improve the diet of Americans, it has to address the key financial barriers to accessing good food, Ortega argued.

“Healthy eating isn’t just a choice. There are real constraints,” he told Fortune following the announcement of the new guidelines. “If nutritional guidance is going to translate into real-world behavior, we have to account for these things: price levels, inflation dynamics, access constraints, and time costs.”

This story was originally featured on Fortune.com

According to the Cengage Group’s 2025 Graduate Employability Report, 48% of graduates say they feel unprepared for entry-level roles. Too often, students are encouraged to “get ready” for their careers. When we reposition “career” as something unfolding now, not next, we invite students to turn inward, and to recognize, examine, and ultimately take ownership of the process of shaping their professional lives.

It’s time to challenge students to rethink the traditional notion of career as a fixed destination, a singular title, or an industry. Career is no longer a static noun but rather a dynamic process, a dialogue between who we are and how we contribute to the world around us. It evolves with each barrier we overcome, each win we celebrate, each identity we carry, and each expectation — both others’ and our own — we learn to navigate. Career isn’t something we reach. It’s something that continually unfolds. It’s not merely what we do; it’s who we are becoming.

Saying that students can have multiple careers isn’t enough. This framing still treats career as something external, implying a certain “otherness” that is separate from the present moment. We should instead help students to see career as work they are already doing, shaped by their choices, values, and growth. What will help students succeed in the early career transition is not contingent on the perfect interview, cover letter, or LinkedIn profile. It comes down to character, self-knowledge, and trust — and the ability to leverage these qualities to navigate complex workplace ambiguity.

So how do we help students adopt this reframed concept? Colleges and universities often focus on transforming students to become career–ready. But the truth is, the transformation starts long before they ever set foot on campus and is ongoing — it happens every day, in ways we often overlook. Career readiness isn’t confined to a classroom or résumé workshop; it’s built through lived experience. It’s in the teamwork and leadership shown on the field, the communication and time management practiced during group projects, and even the professionalism and conflict resolution learned while navigating roommate challenges. Faculty can help reinforce this reframing by connecting course material to professional practice, making clear that the classroom isn’t separate from the career journey, but central to it.

With today’s workforce defined by constant change, we must rethink what career readiness truly means. It’s not preparation for a distant future; instead, it’s the practice of self-awareness and adaptability that begins long before formal job titles. We’ve heard it loud and clear: Students and families are asking for more from the college experience. It’s not enough simply to be career-ready for entry-level jobs; we need students to become career-savvy once they enter the workforce. Students are voicing that traditional career preparation is not enough, “because readiness isn’t just about getting in the door. It’s about what happens after,” as Karen McCullough and Laura Nicole Miller point out in examining the student perspective on career readiness.

The University of New England’s recent Industry Exploration Day reimagines traditional career preparation by embedding experiential learning into career development programming and normalizing learning as part of professional practice. Beyond the familiar career-fair format, students met with employers to workshop their résumés, refine their LinkedIn profiles, and practice interview questions. A student photographer provided professional headshots, building a portfolio while supporting their peers. Redesigning career programming as more than events students simply attend — but as spaces where they actively shape their professional identities — shows that career is a continually unfolding process, one that invites them to step into multiple roles as peers, future colleagues, and emerging mentors.

To truly prepare students for the world of work, we must move beyond teaching them to anticipate it; instead, we need to help them experience it. By simulating real-world challenges, we push students to think on the fly, to read a room, to connect context with action, and to wield the most powerful tool they possess: themselves. Helping students navigate ambiguity means cultivating adaptive expertise — the skill of drawing from what they’ve learned, who they are, and what they value to make sense of what comes next.

But this approach to experiential education only works when we stop treating career as a future destination and start seeing it as a process already in motion. Anticipating change, making decisions with intention, and drawing from lived experience — this is how we prepare students for what comes next.

Career readiness isn’t a path to the future waiting to be discovered; it’s an internal process that’s already well underway. And it’s time higher education acted like it.

The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.

This story was originally featured on Fortune.com

Good morning. You’re not imagining it: CEO turnover is up, and 41% of the 22 incoming CEOs in the S&P 500 this quarter had prior experience running public companies, up from 25% in Q1 last year. And the trend of hiring experienced insiders, which we recently noted in this column, is validated in Russell Reynolds’ latest Global CEO Turnover Index. The CEOs who left were also in those roles longer—11.9 years, on average, vs. 8.3 years last year—and it’s taking longer to fill those roles. Similar patterns are playing out globally. For more insight, I spoke with Russell Reynolds’ CEO Constantine Alexandrakis about the type of leader that is in demand right now:

Leaders who can hit the ground running. “One quarter does not tell a story, but there is more urgency in the air, more speed, more change, more transformation. And an experienced pair of hands at a moment like this could be more attractive,” he told me. Boards want leaders who don’t need a steep learning curve or lengthy onboarding, making insiders and experienced CEOs attractive, especially if they previously sat on the company board.  

They can create the right team. “It’s not about one person. It’s about a system of people and a leadership team that comes together and drives the change. There’s a risk that boards, and the world in general, are over-indexing on the CEO as the one who is going to make all this happen. It’s the system of the leadership team and the collective within the company.”

They focus on people, not technology. “AI is not about technology. It’s about changing behavior and driving change management, which is really difficult to make happen … Sometimes, just a change is needed. Understand what needs to change in order for the transformation to succeed.”

Alexandrakis is trying to follow his own advice. In September, his term as CEO was renewed through December 2030, and he is approaching the job with a fresh perspective. As he put it: “I am being much more of an enabler of a team and a network and a collective—being the center of that group versus being the top of the pyramid. That’s my aspiration.”

Contact CEO Daily via Diane Brady at diane.brady@fortune.com

This story was originally featured on Fortune.com

They all got their start in newspapers, either pedaling routes in the dark, tossing the latest newspaper on the porch, or chasing down customers for payment.

Warren Buffett made some of his first cash by slinging The Washington Post. Tim Cook woke up at 3 a.m. to deliver the Mobile Press Register in Alabama. It taught these future executives some of the values they took all the way to the C-suite.

The paper route teaches “just good, basic business principles,” Ross Perot told the Associated Press in 1995, listing skills like managing inventory, collecting payments, and showing up on time, every day, no matter the weather. The 1992 presidential candidate said he started delivering papers when he was about 12 years old and threw them from his horse in a sand-strewn neighborhood.

The paper route is a relic. Falling print circulation and child labor concerns have handed the job to adults and the Postal Service. But the executives who once had the route haven’t forgotten it—and many say it’s where they learned everything that counts.

Here are five Fortune 500 executives—plus a few honorable mentions—who got their start slinging papers.

Warren Buffett, Berkshire Hathaway

The Oracle of Omaha started delivering The Washington Post and the Washington Star at 13. By 14, he had multiple routes and was earning $175 a month—more than some of his teachers, according to Alice Schroeder’s Buffett biography, The Snowball: Warren Buffett and the Business of Life.

Buffett, 95, filed his first tax return that year, deducting his bicycle and his watch as business expenses, he told PBS News Hour in 2017. The job, he has said, taught him lessons that aged well.

“You learn a lot about human nature when you deliver papers,” Buffett said. “For one thing, you learn you have to pay for them each month. Whether the customers pay you or not. You have to collect money.”

Buffett was once one of the largest shareholders of the paper he once tossed on doorsteps—and said he even considered buying it when it went up for sale in 2013. Amazon founder Jeff Bezos now owns The Washington Post. Buffett retired as Berkshire Hathaway CEO at the end of 2025 (he still serves as chairman). He’s worth an estimated $141 billion, according to the Bloomberg Billionaires Index.

Tim Cook, Apple

Apple’s outgoing CEO (to be replaced by John Ternus later this year) got his first job at 11 or 12, delivering the Mobile Press Register in Robertsdale, Ala. The paper stopped print publication in 2023.

“Throwing papers helped start my college education,” Cook, 65, told The Wall Street Journal of becoming the first member of his family to attend college. After delivering papers, Cook said he “graduated to flipping burgers” at a local Tastee Freeze for $1.10 an hour, he told the Table Manners podcast. Cook is worth an estimated $3 billion, according to Forbes.

Michael Dell, Dell Technologies

The future PC mogul sold subscriptions to the now-defunct Houston Post as a teenager, and quietly invented the playbook that would become Dell. 

Instead of cold-calling, he pulled marriage licenses and new-mover records and sent direct mail to the most likely buyers, earning $18,000 in a year. 

“It was an early lesson in direct marketing, for sure,” Dell said at SXSW in 2024.

Walt Disney, the Walt Disney Co.

The animator-to-be and the man would go on to build one of the world’s largest entertainment companies, delivering the Kansas City Star and the Kansas City Times with his brother Roy, starting at age 9. 

“When I was 9, my brother Roy and I were already businessmen,” Disney said, according to a Disney archive. “We had a newspaper route…delivering papers in a residence area every morning and evening of the year, rain, shine, or snow.”

They woke up at 3:30 a.m., worked until it was time for school, and “did the same thing again from four o’clock in the afternoon until supper time.” 

Walt Disney died in 1966 and, at the time, had a net worth estimated at roughly $100 million to $150 million, which would be about $1 billion to $1.5 billion today. He was the founder and CEO of what became a Fortune 500 company.

Ross Perot, Electronic Data Systems

The Texas billionaire and two-time presidential candidate may have had the most cinematic route of the bunch. As a boy in Texarkana, Perot built a Texarkana Gazette route from scratch in a part of town the paper had largely ignored—and delivered it on horseback, traveling 20 miles a day, he told Fortune in 1968.

Because he had launched the territory himself, he negotiated 70% of the subscription price instead of the standard 30%, and fought the paper when it later tried to claw the cut back, he told Fortune.

In 1962, Perot founded Electronic Data Systems, which was a Fortune 500 company. HP bought EDS in 2008 in a $13 billion deal. Perot died in 2019, and was worth about $4 billion.

Honorable mentions: garbage bags, graveyard shifts, and Big Macs

Not every CEO started by delivering the newspaper. 

Mark Cuban got his start at 12 selling garbage bags door-to-door to pay for basketball shoes, and Jeff Bezos worked the breakfast shift as a short-order cook at McDonald’s in high school.

Former PepsiCo CEO Indra Nooyi worked the midnight-to-5 a.m. receptionist shift at her Yale dorm because it paid 50 cents more an hour, and GM CEO Mary Barra got her start at the automaker itself, as an 18-year-old co-op student inspecting fender panels on the assembly line.

While they all had different first jobs, they all shared similar work-life lessons: show up, be on time, and be on time every morning, even when there’s three feet of snow at 4 a.m.

This story was originally featured on Fortune.com

AI isn’t destroying jobs. It’s making workers dramatically more productive. And the workers are doing that extra production? They have no idea.

The numbers that should calm everyone down

The Morgan Stanley report, authored by Chief U.S. Economist Michael Gapen and a team of economists, examined industry-level output per employee across the U.S. economy and cross-referenced it with each industry’s degree of AI exposure. The results were striking: industries classified in the top quartile of AI exposure contributed 1.7 percentage points to the overall 2.4 percentage-point growth in productivity recorded over the four quarters through the end of 2025. A year earlier, those same industries had contributed just 0.7 percentage points. The acceleration is not subtle.

Here’s what makes the finding particularly revealing: that surge in productivity wasn’t produced by cutting headcounts. Employment trends across high-, medium-, and low-AI industries were broadly similar. What differed was output — how much those workers were producing. In high-AI industries, output accelerated sharply while employment growth stagnated. In low-AI industries, output actually slowed.

In economic terms, this appears to be a best-case scenario unfolding in real time: workers are not being displaced; they are being augmented. But psychologically and culturally, it creates a paradox. The workforce’s hatred of AI may look, from the outside, like a failure of economic literacy. Workers are thriving and don’t know it. Look closer, and a more sophisticated fear comes into view — one that the productivity data doesn’t address at all.

The 90% problem

But aggregate productivity numbers obscure a brutal internal sorting that is already underway inside companies. Tech executive and AI strategist Daniel Miessler, whose observations on the workforce have circulated widely in recent weeks, argued in a LinkedIn post that the real dynamic isn’t AI replacing workers — it’s AI allowing a small tier of top performers to absorb the work of everyone below them.

“AI can’t come anywhere close to replacing the top performers at a big company,” Miessler wrote. “But they’re spending millions a year on tens of thousands of employees in the bottom 75%… companies no longer want to pay millions a year for mediocre employees. They’d rather fire everyone but the best, and have them become 10x or 100x what they were by wielding AI.” The productivity boom, in this reading, isn’t lifting all boats. It’s concentrating leverage at the top while quietly marking a much larger cohort for displacement — not by machines directly, but by a smaller number of humans wielding them. An even darker outcome is whether those top 10% workers are just buying themselves a few more years before they’re displaced, too.

The AI tools driving this boom, technologist Shaun Warman pointed out in a recent blog post, are not priced at their actual cost. A serious individual user of a frontier model consumes roughly $80 to $150 of compute per month at real prices; the subscription that buys it runs $20. OpenAI has acknowledged publicly that even its $200-a-month enterprise tier loses money on its heaviest users. The reason for the subsidy, according to Warman, is simple and unsettling: “The user is not yet the customer. The user is the training set.” Every edit, regeneration, and follow-up question a worker fires into a frontier model is training data, aggregated across hundreds of millions of users and tens of billions of conversations.

“Synthetic data has crossed the quality threshold,” Warman argued. “Models can now generate, filter, and grade their own training data at a level competitive with raw human input. The frontier labs publish papers on this monthly.” The straightforward implication, he concluded, is that “the marginal value of a human edit is falling as the model’s ability to produce its own corrections rises.”

What happens when the subsidy ends?

Warman identified three forces that will close what he calls “the apprenticeship window” within three to five years: the aforementioned quality threshold; agentic self-play, allowing models to evaluate and improve themselves in domains with verifiable outcomes; and sheer scale, meaning that additional human feedback is hitting diminishing returns. When those forces converge, the subsidy that makes AI accessible to ordinary workers will lose its justification.

Warman’s predictions for what follows are stark: the $20 monthly tier will vanish or degrade into an advertising-supported shadow product; top capabilities get gated behind enterprise contracts with five-figure annual minimums; and in the most extreme scenario, the labs stop licensing AI as a tool altogether and simply become the operator — law firms, consulting shops, hedge funds — capturing the economic value themselves rather than sharing it.

The productivity data, in other words, is a snapshot of the apprenticeship phase — a period when workers’ and labs’ interests align almost perfectly. Both want better output. For now, both get it cheaply.

The Morgan Stanley economists note, carefully, that the pattern of rising output and stable employment “may change as AI adoption picks up.” That hedge deserves more weight than it typically receives — because the repricing, when it arrives, will not affect all workers equally. Large enterprises that can amortize six-figure AI contracts across thousands of employees will barely notice the transition. Workers at smaller firms, in lower-margin industries, or in the public sector — teachers, healthcare aides, municipal workers — face a starker calculus: the tools that made them more productive were priced below cost as a temporary subsidy, and the bill, when it comes, will be set by what large corporations are willing to pay, not what individual workers can absorb. The productivity gains currently showing up in the data may accrue broadly. The tools that generated them may no longer be broadly accessible.

Workers aren’t wrong to be wary of AI. They may just be sensing the right threat at the wrong moment — not the displacement that’s being measured, but the repricing that isn’t yet.

For this story, Fortune journalists used generative AI as a research tool. An editor verified the accuracy of the information before publishing.

This story was originally featured on Fortune.com