The telecom giant is in the midst of a multi-year transformation toward a leaner, AI-driven model. On the company’s Q1 earnings call Monday, Dan Schulman, chief executive since October, pointed to churn as “the clearest measure” of whether the company’s efforts are resonating.

Verizon, No. 30 on the Fortune 500, reported its first positive first-quarter postpaid phone net adds since 2013—a net addition of 55,000 postpaid phone subscribers. Postpaid customers, who pay monthly bills under contract, are considered the most valuable because they carry the largest bills and are less likely to switch providers.

Consumer postpaid phone churn was 90 basis points in Q1, a sequential improvement of 5 basis points from Q4, and improved further in March to below 85 basis points. “That is a significant improvement both sequentially from Q4 and within the quarter, and it reversed the upward pressure we had seen in churn over the past several years,” Schulman said.

Lowering churn makes Verizon’s “marketing dollars work harder because we are not simply replacing customers who leave; we are adding to a more stable base,” he said. The company is no longer predominantly reliant on expensive promotions to drive growth, instead focusing on a “disciplined, repeatable, and fiscally responsible” approach, he added.

Schulman, the former CEO of PayPal and previously Verizon’s lead independent director, succeeded Hans Vestberg. He was charged with steering Verizon’s next phase of customer focus and financial growth. Under Vestberg, Verizon struggled to articulate a clear strategy around market positioning, branding, and pricing, according to analysts.

Customer metrics take center stage

Schulman highlighted customer-centric metrics: churn, customer acquisition, and customer lifetime value (CLV). These metrics have evolved from marketing roots to become pivotal in finance and the bottom line.

“Our cost of acquisition [CAC] and retention in March was down approximately 35% relative to the end of Q4, and we expect to maintain a lower cost of acquisition and retention as we look forward,” Schulman said. “These trends in churn and unit economics are lifting our consumer lifetime value and are already flowing through to the bottom line and to our free cash flow,” he added.

Morningstar Equity Director Michael Hodel wrote in an analyst note Monday that although Verizon’s wireless customer churn remains elevated, reflecting the competitive environment, the company is “doing a much better job attracting new customers than a year ago.” Lower pricing has helped drive customer acquisition, Hodel said. Morningstar maintained its $53 fair value estimate and narrow economic moat rating.

Telecom CFOs track postpaid net adds, churn, CAC, and CLV as key drivers of revenue guidance and capital allocation. Verizon still lost more postpaid phone subscribers than it gained in the quarter, but only by 35,000, a 321,000 improvement from a year ago, CFO Tony Skiadas said on the call. The gains were driven by a healthier mix of genuinely new customers coming in and fewer existing ones walking out. “While there is more work to be done with customer experience, which is the largest component of our transformation plan, we’re pleased to see early signs of progress towards our goals,” he said.

Total operating revenue for Q1 was $34.4 billion, up 2.9% year-over-year. Adjusted EPS was $1.28, beating estimates, up 7.6% year-over-year, which is the highest growth rate in over four years. Consolidated net income was $5.1 billion, up 3.3%. Verizon raised its 2026 adjusted EPS growth forecast to 5%–6%. Its stock price gained around 1.5% at the close Monday, but it had risen as much as 4.5% earlier in the trading session. 

Schulman noted that there is still work ahead. “We entered 2026 with a clear set of priorities, a step function improvement in guidance, and a realistic plan,” he said. 

Sheryl Estrada
sheryl.estrada@fortune.com

This story was originally featured on Fortune.com

But innovation, creativity, and entrepreneurial acumen aren’t restricted by borders, as I was reminded when talking to Joaquín Cuenca Abela recently. The 49-year-old Spanish founder is showing how it’s possible to thrive in the AI market even if your company isn’t building a frontier model, even if it’s not backed by VC money—and even if it isn’t based in Silicon Valley.

Back in 2010, a few years after selling his startup to Google, Cuenca cofounded a company called Freepik in Málaga—the sun-drenched birthplace of Pablo Picasso on Spain’s southeastern coast. Freepik carved a profitable niche for itself as one of the most popular online platforms for stock images, and the company helped establish Málaga as an emerging tech hub that has attracted companies like Google, Oracle, and Vodafone. 

When OpenAI’s DALL-E 2 image generator came out in 2022 though, Cuenca realized everything was about to change and he pivoted hard into generative AI. Freepik began offering tools that combined AI image-generating models with editing tools. Last year, he pivoted even further and pushed the company into AI video generation. 

Today, the business is generating $230 million in annual recurring revenue, with video accounting for roughly half of the revenue, Cuenca tells Fortune exclusively. And given that its business has changed so much from its initial days as a stock imaging platform, the company is changing its name from Freepik to Magnific.

“We are creating a new economy,” says Cuenca, who is the CEO. “It’s not that we are getting users from any particular competitor, it’s that people are finding that they can do a new thing that was not possible before.”

Magnific isn’t trying to compete with the big model makers. It lets users pick from various video AI models, including Google’s Veo 3.1 and ByteDance’s Seeddance 2.0, and combines it with its own tools. Magnific’s pre-production tools, for example, let users create assets like images of characters, props, and scenery to create a polished AI video that’s consistent with the storyline. Magnific’s product has been used in ad campaigns for Puma and Carl’s Jr, as well as in the Amazon Prime Video series House of David, and by the BBC, among others.

Making the pivot from an established business to something new like video AI wasn’t easy, Cuenca acknowledges. The company currently has 400 employees, down from roughly 550 when it was focused on stock images. “There’s going to be some pain in any transformation,” Cuenca says, but notes the company is hiring employees with different skills now and that the intention is to ultimately get bigger than Freepik was at its peak. The company opened an office in San Francisco in 2023 that currently has around 20 employees, and also has an office in Colombia. 

Throughout the journey from Freepik to Magnific, Cuenca says the company has been bootstrapped the entire time. He’s never raised money from outside investors, and he says Magnific is profitable. 

Would he consider raising money in the future, especially given the not-so-trivial costs of generating tokens for AI? At some point, perhaps, Cuenca says. But, “if we do it, it’s because we want to grow the DNA of the company.”

See you tomorrow,

Alexei Oreskovic
X: @lexnfx
Email: alexei.oreskovic@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

He noted that he would have $15 trillion “left over” if he managed to sell 10 million cards, adding: “It may be earmarked for deficit reduction, but it actually could be more money than that.”

Trump encouraged reporters to “remember the words ‘gold card'” last year, and on Thursday, Commerce Secretary Howard Lutnick had an update: One person has been approved. “There are hundreds in the queue that they are going through,” Lutnick added to a congressional committee hearing last week.

Despite promising a year ago that gold cards would raise $1 trillion in revenue, Lutnick updated the committee that the “set up” of the scheme has now been completed, and the team “wanted to make sure they did it perfectly.”

While budget hawks will welcome any revenues earmarked for debt reduction, there are some questions over the feasibility of the scheme: More importantly, how many immigrants are wealthy enough to fork out $5 million per card, or $20 million for a family of four.

Therein lies the problem: Knight Frank’s Wealth Report for 2026, published last week, broke down where the globe’s ultra-high net worth individuals (UHNWs) live—defined as people with more than $30m in assets.

While there has been significant growth in regions such as Indonesia, Saudi Arabia and Poland, the UHNW populations of each are still relatively small. For example, the Middle East as a whole is home to just 3% of the UHNW population.

Europe as a whole, already home to pockets of wealth in the likes of London and Paris, is home to 22.7% of the world’s UHNW population. North America, on the other hand, is home to 42.6% of the population of the world’s wealthiest.

In other words, the predominant portion of people able to afford gold cards are likely to already live in the U.S. as it is, by far, home to the most millionaires.

Tariff question

Trump had also proposed tariffs as another revenue stream for helping to pay down national debt.

The duties have indeed proved a major boon to the balance sheet, the Yale budget lab reporting earlier this month that tariffs collected in 2025 raised an estimated $214.7 billion in inflation-adjusted customs revenue above the 2022–2024 average, roughly around $300 billion a year.

The method has proved so effective that economists widely expect future administrations to keep the levies in place—despite their unpopularity with foreign trade allies. But under Trump, how the proceeds will be used remains a question.

For example, the president previously said he would be sharing the funds in the form of a $2,000 rebate checks. However, calculations by Fortune found that even if the government were to pay each household—as opposed to each individual—in the bottom 50% of earners, that would still require payouts to more than 67.5 million homes, per data from the St. Louis Fed. That would imply $135 billion, or nearly half of the tariff revenues, would immediately be paid out to citizens instead of being directed toward national debt.

There’s also the question of the One Big Beautiful Bill Act. The Committee for a Responsible Federal Budget highlighted last month that, taking into account its dynamic effect on the economy, spending and tax breaks in the act will add $4.2 trillion to the national debt through Fiscal Year (FY) 2034 or $4.7 trillion through 2035.

Previously, much of this had been earmarked to be offset by tariffs—which had been estimated to reduce deficits by $2.5 to $3 trillion through 2035. However, in the wake of a Supreme Court ruling earlier this year about the legal basis for 2025 tariffs, the Congressional Budget Office (CBO) has reexamined its outlook. In March, the CBO wrote that the termination of the tariffs resulted in deficit increases of $2 trillion greater by 2036 than previously projected.

This story was originally featured on Fortune.com

Hyperliquid is a decentralised exchange that processes billions of dollars in daily trading volume across perpetual derivative  and spot markets, allowing users to trade an expanding range of digital and real-world assets including stocks, crude oil, silver, and equity indices. It operates on a purpose-built blockchain whose execution engine and consensus mechanism are optimized for trading efficiency—trades settle in under a second, compared with the T+1 window still standard in equities, with no downtime and transaction costs that are a fraction of what traditional venues charge. Hyperliquid does this without any traditional centralised intermediaries, exchange operators, or clearinghouses and on a 24/7 basis without market closures.

For most of the past decade, digital assets have been discussed in two narrow frames: speculation and store of value. Bitcoin was digital gold; everything else was a casino. That framing was always incomplete, but it was understandable given the immaturity of the technology and the absence of real economic activity on-chain.

What is emerging now looks different. In late February, when the United States and Israel struck Iran on a Friday evening NY time, traditional commodity derivatives markets were closed. Trading in perpetual derivatives in crude oil on Hyperliquid surged within hours as participants priced in the shock in real time. By the weekend, trading in oil perpetuals on the platform recorded over $1.2 billion in 24-hour notional volume. As major Western commodity derivatives markets reopened on Monday, they confirmed the direction Hyperliquid had already been pricing in for two days. Silver followed a similar pattern in January during a period of acute physical market stress. In March, S&P 500 perpetuals began trading on the platform with the full endorsement from S&P Dow Jones Indices, offering round-the-clock, on-chain exposure to the world’s most widely tracked equity index.

These are not toy markets. This is genuine price discovery happening on decentralised infrastructure 24/7 and it is happening precisely because the infrastructure is available when and where legacy venues are not.

The economics merit attention too. Hyperliquid generated approximately $962m in fees in 2025 on roughly $3tn in notional trading volume. The majority of these fees flow into an automated, non-discretionary mechanism that purchases HYPE, the network’s native token, on the open market, reducing the circulating supply of the token. HYPE secures the network through staking, provides trading fee discounts, and conveys governance rights—all of which promote use and adoption of Hyperliquid. Unlike purely speculative tokens, its economics are a product of the protocol’s design: trading activity generates fees, fees reduce supply, and growing adoption reinforces the network’s security and liquidity.

This matters because of what it signals about market structure. When we invested in Circle in 2021, our thesis was about infrastructure, not cryptocurrency. Circle was building a digital version of the dollar for institutions that could move at the speed of the internet, settle around the clock, and eliminate layers of intermediaries. The same logic applies here, at the trading layer rather than the payments layer. Blockchain technology is not replacing the financial system. It is rewiring it.

None of this is to suggest the regulatory questions are settled or straightforward. Hyperliquid sits outside the existing U.S. market structure framework, and as a permissionless, non-custodial exchange it raises genuine questions across several regulatory perimeters: what is the jurisdictional line between SEC and CFTC oversight? What does anti-money laundering and sanctions compliance look like in an architecture without a single identifiable operator? How to best address the cross-border problem of whose rules apply when validators, liquidity providers, and end users are globally distributed? And, when there is no central counterparty in the traditional sense, where does enforcement fall? Should it be on front-ends, on market makers, on validators, or on the protocol itself? How these questions are answered will shape Hyperliquid’s trajectory and the development, access to and adoption of next generation market infrastructure in the United States. Recent announcements and enforcement patterns suggest US regulators are aware of the challenges and are working hard on constructive answers.

For too long, the conversation about digital assets has been trapped between price speculation and regulatory anxiety. Both miss the point. What is being built on platforms like Hyperliquid is not a speculative playground. It is the next generation of market infrastructure, which is open, efficient, and available to anyone with an internet connection. The settlement window, in every sense, is closing.

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

The answer can have serious implications in a variety of scenarios, including when an employee is owed overtime pay and whether an additional company is obligated to pay it. It also determines whether an additional company is responsible for any assessed penalties if they violate federal labor law by having a young employee perform dangerous labor.

That’s why the United States Department of Labor is taking steps to provide greater clarity. This week, we published a proposed rule — the Joint Employer Status Under the Fair Labor Standards Act, Family and Medical Leave Act, and Migrant and Seasonal Agricultural Worker Protection Act — designed to clear up confusion and demystify joint employment for both employers and employees.

Published in the U.S. Federal Register on April 22, our proposed rule seeks to eliminate confusion and uncertainty surrounding when joint employment applies and to clarify the scope of joint employment in the regulatory space. It also aims to improve labor law compliance by defining established roles for employers and employees. Furthermore, the proposal would simplify the department’s enforcement efforts, reduce burdensome litigation, and promote greater uniformity in court analyses.

What Joint Employment Actually Means

Joint employment occurs when two or more employers share legal responsibility for the same employee and is especially common in industries like staffing, franchising, construction, and farming. When an employee is subject to joint employment, all employers involved are jointly liable for ensuring the employee receives the correct wages and benefits, including overtime.

The New “Bonnette” Test

To provide greater clarity, the proposed rule includes distinct standards for different joint employment situations. The most important of these is “vertical” joint employment in which one company employs a worker whose labor may also benefit another company. This leads to a dilemma: should the other company be considered an additional employer?

To address this situation, the department is proposing a modified version of the “Bonnette” test — something that would help employers and the department’s investigators determine whether joint employment applies.

Under the proposed standard, joint employment would be determined by assessing whether the potential joint employer: hires or fires the employee; supervises and controls the employee’s work schedule or employment conditions to a substantial degree; determines the employee’s rate and method of payment; and maintains the employee’s employment records.

These factors are the closest we have to a common denominator applied by courts when determining joint employment under the Fair Labor Standards Act. By synthesizing these criteria, the department aims to encourage greater consistency and uniformity for stakeholders — including employees themselves.

What the Rule Rules Out

In addition to the test, we’ve introduced guidance that parties generally should not consider certain criteria that are confusing, irrelevant, or even contradictory. Some courts, for example, have considered extraneous factors that should only matter in understanding whether a worker is an employee or independent contractor — an entirely separate issue. The proposal also provides that certain common business models and practices, such as providing a sample employee handbook or requiring anti-harassment training, don’t by themselves make joint employer status more likely.

By providing clear and uniform standards through this proposed rule, the Department of Labor anticipates less time and money will be spent trying to determine whether joint employment applies. We’re also confident it will help employers avoid labor law violations.

How to Weigh In

Releasing this proposal is just the first step in the rule-making process. In the weeks to come, I invite all parties to provide public comments to the department during the 60-day comment period — open until June 22 at regulations.gov. With your support — and under President Trump’s strong leadership — the Department of Labor will continue to put American workers first each and every day.

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

• OpenAI missed its targets for revenue.
• Inside Meta’s Manus fiasco: “Killing the chicken to scare the monkeys.”
• AI capex will hit $660 billion this year.
• Markets: New record highs but oil is $111.
• 🚢 A tanker escaped from the Strait of Hormuz.
• Moving house is now a “luxury good.”

This story was originally featured on Fortune.com

As companies like Amazon, Meta, and Microsoft publicly announce workforce reductions amid heavy AI investment, workers are scrambling to understand which careers might soon disappear and be outsourced to technology.

A report from Microsoft researchers studying the occupational implications of generative AI offers some clarity.

Translators, historians, and writers are among the roles with the highest AI applicability score, meaning the job’s tasks are most closely aligned with AI’s current abilities, according to the 2025 report that ranked professions. Customer service and sales representatives—which make up about 5 million jobs in the U.S.—will also have to compete with AI. 

Overall, the jobs most exposed are ones that involve knowledge work—like people doing computer, math, or administrative work in an office, the researchers wrote. Sales jobs are also high on the list, since they often involve sharing and explaining information.

While Microsoft said high applicability doesn’t automatically mean those jobs will necessarily be replaced by AI, the list of roles quickly went viral—with professionals deeming them “most at risk.” It comes as companies have been freezing thousands of would-be new roles that it expects AI will take over in the next 5 years, and graduates in the U.K. are facing the worst job market since 2018 as employers pause hiring and use AI to cut costs, according to Indeed.

Of course, there are some jobs that are unlikely to be touched by AI: Dredge operators; bridge and lock tenders; and water treatment plant and system operators are among the jobs with virtually no generative AI exposure, thanks in part to their hands-on equipment requirements.

Still, business leaders like Nvidia CEO Jensen Huang have said every job will be touched by AI in some way, and so it’s best to embrace it. 

“Every job will be affected, and immediately. It is unquestionable,” Huang said at the Milken Institute’s Global Conference in 2025. “You’re not going to lose your job to an AI, but you’re going to lose your job to someone who uses AI.”

A degree won’t save you from AI’s jobs revolution

Many of the jobs with high chances of getting upended by AI soon, like political scientists, journalists, and management analysts, are all ones that typically require a four-year degree to land a job. And as the researchers point out, having a degree—which was once considered a surefire path to career advancement—is no longer a safeguard against the changing tides. 

“In terms of education requirements, we find higher AI applicability for occupations requiring a Bachelor’s degree than occupations with lower requirements,” wrote the researchers, who studied 200,000 real-world conversations of Copilot users and cross-compared the AI’s performance with occupational data.

On the flip side, there are some career paths with low AI exposure, that are growing in demand. The health care sector, in particular, is an area that is experiencing this heavily. The home health and personal care aid industry is expected to create among the greatest number of new jobs over the next decade, according to the U.S. Bureau of Labor.

At the same time, the researchers recognized even their findings don’t capture the full scope of the AI revolution—and there could be further automation caused by more than just generative technology: “Our measurement is purely about LLMs: other applications of AI could certainly affect occupations involving operating and monitoring machinery, such as truck driving.”

Kiran Tomlinson, a senior Microsoft researcher, told Fortune the study focused on highlighting where AI might change how work is done, not take away or replace jobs.

“Our research shows that AI supports many tasks, particularly those involving research, writing, and communication, but does not indicate it can fully perform any single occupation. As AI adoption accelerates, it’s important that we continue to study and better understand its societal and economic impact,” Tomlinson said.

Gen Z’s big bet on education might not be all glam

After seeing the rollercoaster of layoffs across the tech industry over the past few years, many Gen Zers have turned to seemingly steadier fields like education.

The sector was the fastest-growing industry among recent U.K. graduates last year, and it was similarly a top career choice for American graduates. And while the profession can provide further work-life balance and decent benefits, the ability for AI to do the work may cause further headache. The report singles out farm and home management educators—as well as postsecondary economics, business, and library science teachers—as roles with relatively high AI applicability.

While it’s unlikely that schools will roll out AI teachers en masse, the report’s findings underscore how quickly the technology could reshape the education profession—and many others.

The top 10 least affected occupations by generative AI:

1. Dredge Operators
2. Bridge and Lock Tenders
3. Water Treatment Plant and System Operators
4. Foundry Mold and Coremakers
5. Rail-Track Laying and Maintenance Equipment Operators
6. Pile Driver Operators
7. Floor Sanders and Finishers
8. Orderlies
9. Motorboat Operators
10. Logging Equipment Operators

The top 40 most affected occupations by generative AI:

1. Interpreters and Translators
2. Historians
3. Passenger Attendants
4. Sales Representatives of Services
5. Writers and Authors
6. Customer Service Representatives
7. CNC Tool Programmers
8. Telephone Operators
9. Ticket Agents and Travel Clerks
10. Broadcast Announcers and Radio DJs
11. Brokerage Clerks
12. Farm and Home Management Educators
13. Telemarketers
14. Concierges
15. Political Scientists
16. News Analysts, Reporters, Journalists
17. Mathematicians
18. Technical Writers
19. Proofreaders and Copy Markers
20. Hosts and Hostesses
21. Editors
22. Business Teachers, Postsecondary
23. Public Relations Specialists
24. Demonstrators and Product Promoters
25. Advertising Sales Agents
26. New Accounts Clerks
27. Statistical Assistants
28. Counter and Rental Clerks
29. Data Scientists
30. Personal Financial Advisors
31. Archivists
32. Economics Teachers, Postsecondary
33. Web Developers
34. Management Analysts
35. Geographers
36. Models
37. Market Research Analysts
38. Public Safety Telecommunicators
39. Switchboard Operators
40. Library Science Teachers, Postsecondary

A version of this story originally published on Fortune.com on July 31, 2025.

More on artificial intelligence:

• Fewer than 1 in 4 workers feel their job is safe. Here’s why worker ‘FOBO’—fear of becoming obsolete—is hurting companies
• Nvidia CEO Jensen Huang says you won’t lose your job to AI—you’ll lose it to your coworker who uses it
• AI is cutting 16,000 U.S. jobs a month—and Gen Z is taking the brunt, Goldman Sachs says

This story was originally featured on Fortune.com

When her and Khloe Kardashian’s Good American denim line dropped, it made $1 million on day one, making it the biggest denim launch in apparel history. Grede has helped redefine inclusion in retail and became the first Black female investor on Shark Tank—all before turning 45.

But for all her business instincts, Grede admits there’s one area where she needed a push—and fellow ex-Shark Tank star Mark Cuban was the one who gave it to her.

In an exclusive 2025 interview with Fortune, Grede talked about an episode of her hit podcast show Aspire that hadn’t aired yet, where the two sat down and compared their AI usage. 

“I was already kind of getting there, but if I’m really honest, that episode where we really delved into AI gave me a new urgency around how I use AI,” she recalled, adding that Cuban had a staggering 60 AI apps on his phone. “Yeah, he gave me a kick.”

As soon as the recording wrapped up, she said that she started looking into AI courses at the Wharton School and Harvard for this fall. “I need to figure this out, because I’m using AI like a 42-year-old woman,” Grede candidly admitted while laughing.

Grede gave her staff a cash bonus for using AI—long before she realized she was behind

Grede isn’t completely new to AI. In fact, she was ahead of the curve when it came to encouraging AI adoption within her companies.

“About two years ago I put a note out in my office giving a cash bonus to anyone that uses AI in their work,” she explained, adding that the incentive was a big hit—especially with the marketing and finance teams.

“It changed the office. It changed the way people presented their work. It changed the way people did their work.”  

But as her Gen Z and thirtysomething staff embraced experimenting with ChatGPT and other new AI launches at the time, Grede admitted that she perhaps leaned too heavily on them. It meant that until recently, she’s been using AI more as a search engine and leaving her staff to handle the rest.

“I was like, Emma, you need to sort that out.”

It comes as many other CEOs are scrambling to appoint AI leaders, future-proof their business, and brace for change.

Billionaire Microsoft cofounder Bill Gates says AI is moving at a speed that “surprises” even him and that even if workers learn how to use the latest tech tools, they may still find themselves out of a job. Meanwhile, an ex-Google exec says CEOs are currently too busy “celebrating” their efficiency gains to see they’re next on AI’s chopping board.

But she’s not using AI to make her more productive

As one of America’s richest self-made women—with a reported net worth of nearly $400 million and at least four major businesses to her name—Grede is clearly ruthlessly efficient. But in her eyes, AI isn’t about squeezing even more productivity from her day.

“I’m probably the most productive person in the world. I don’t know that I can be that more productive,” she said, noting that her time is mostly spent making high-stakes decisions—not executing tasks. 

“There’s no amount of AI that can help me with that.”

But where it can help, she said, is in making smarter strategic choices and reshaping how she leads. 

“I think it’s a reframing of how we’re going to do things,” Grede added. “So much of my job is about making really big bets and decisions. And so if I can put data in places to optimize that decision making, I think that that’s probably where I’m going to be using it most.

“When you start to think about my role as a merchant and as a planner, it’s really those things that I think are going to fundamentally shift.”

In her podcast episode with Cuban, which has since been released, the television personality and Dallas Mavericks owner had a stark warning for founders who don’t embrace AI: “You’re f–ked…That’s like saying, back in the day, ‘I don’t need to use a PC, I don’t need to use the internet. I don’t need a cell phone.’”

He added, “If you’re an entrepreneur, or want to be an entrepreneur, start playing with it to get a sense for how it works, how to prompt. It becomes like having an entire staff of 1,000 business professors.”

A version of this story originally published on Fortune.com on August 12, 2025.

Read more success interviews from Fortune’s Orianna Rosa Royle:

• Astrid & Miyu’s founder bought two flats in her 20s—one paid for her MBA and the other funded her $45 million-a-year jewelry brand
• Student discounts made him a millionaire, a heart condition made him rethink life—now this millennial founder spends half the year in the French Alps
• Born in Soviet Union, Grindr CEO was told he had two career options: Learn English or how to shoot a gun
• Scale AI’s 30-year-old billionaire cofounder has a warning for anyone who craves work-life balance: ‘maybe you’re not in the right work’
• At 61, this Fortune 500 CEO still works out 6 days a week with his 23-year-old son—he picks the Gen Zer’s brain for perspective while lifting weights

This story was originally featured on Fortune.com

Though Buffett, 95, will still be in attendance this year, but he’s not set to speak, according to the meeting schedule. And even though he’ll be less hands-on this year, Buffett is still chairman of Berkshire’s board of directors and remains the company’s largest shareholder, holding about 30% of the voting interest and 13.7% of the economic interest.

Every year between 1965 and 2024, Buffett wrote a letter to shareholders ahead of the annual “Woodstock for Capitalists.” Here are some of the “Oracle of Omaha’s” best reflections from his annual letters:  

The best holding period? Forever

Coca-Cola and Apple rank among Berkshire’s most successful investments, but when Buffett started to buy Coke shares in the 1980s, the choice wasn’t so obvious.

“We made major purchases of Federal Home Loan Mortgage Pfd. (“Freddie Mac”) and Coca-Cola. We expect to hold these securities for a long time. In fact, when we own portions of outstanding businesses with outstanding managements, our favorite holding period is forever,” Buffett wrote in his 1989 letter.

In the years after that letter, Berkshire bought 400 million shares of Coca-Cola stock, spending about $1.3 billion in total. Now, the company owns 9.3% of Coca-Cola, worth more than $31 billion. 

Don’t be a duck 

In 1998, Buffett warned against overstating your impact. 

“In a bull market, one must avoid the error of the preening duck that quacks boastfully after a torrential rainstorm, thinking that its paddling skills have caused it to rise in the world. A right-thinking duck would instead compare its position after the downpour to that of the other ducks on the pond,” he wrote. 

“So what’s our duck rating for 1997? The table on the facing page shows that though we paddled furiously last year, passive ducks that simply invested in the S&P Index rose almost as fast as we did.”

Where we went wrong in 2008

In 2009, in the wake of the financial crisis, Buffett shared one place he thought Wall Street went wrong. 

“When the financial history of this decade is written, it will surely speak of the Internet bubble of the late 1990s and the housing bubble of the early 2000s. But the U.S. Treasury bond bubble of late 2008 may be regarded as almost equally extraordinary,” he said. He added that investors who “clinging” to cash equivalents and  long-term government bonds felt validated by commentators saying “cash is king” would be in for a rude awakening. 

“Approval, though, is not the goal of investing. In fact, approval is often counter-productive because it sedates the brain and makes it less receptive to new facts or a re-examination of conclusions formed earlier. Beware the investment activity that produces applause; the great moves are usually greeted by yawns,” he wrote. 

Betting on the U.S. 

In one of his last letters, Buffett shared how being a U.S.-based company contributed to Berkshire’s success. 

“At Berkshire we hope and expect to pay much more in taxes during the next decade. We owe the country no less: America’s dynamism has made a huge contribution to whatever success Berkshire has achieved – a contribution Berkshire will always need. We count on the American Tailwind and, though it has been becalmed from time to time, its propelling force has always returned,” Buffett wrote in 2023.  

“I have been investing for 80 years—more than one-third of our country’s lifetime. Despite our citizens’ penchant—almost enthusiasm—for self-criticism and self-doubt, I have yet to see a time when it made sense to make a long-term bet against America. And I doubt very much that any reader of this letter will have a different experience in the future.”

‘The casino now resides in many homes’

In 2024, Buffett offered prescient reflections on the unruliness of the market, long before the prediction markets began to sway Wall Street. 

“Though the stock market is massively larger than it was in our early years, today’s active participants are neither more emotionally stable nor better taught than when I was in school. For whatever reasons, markets now exhibit far more casino-like behavior than they did when I was young. The casino now resides in many homes and daily tempts the occupants.” 

“One fact of financial life should never be forgotten. Wall Street – to use the term in its figurative sense – would like its customers to make money, but what truly causes its denizens’ juices to flow is feverish activity. At such times, whatever foolishness can be marketed will be vigorously marketed – not by everyone but always by someone.”

Learning from the master

Buffett’s successor, Greg Abel, has continued the yearly tradition and wrote his first shareholder letter in February. He began by praising Buffett. 

“Warren Buffett is arguably the greatest investor of all time, with generations benefiting from his

investment acumen,” Abel wrote. “He has also been a remarkable CEO, executing his vision of building a great insurance business since the acquisition of National Indemnity in 1967, and deploying the float to make successful investments across major sectors of the economy, concentrating in the U.S. (To Warren’s great frustration, this letter begins with these observations – yet we all know they are true.)” 

This story was originally featured on Fortune.com

Good morning. What does it mean to be agile during these volatile, fast-moving and often dangerous times? Fortune co-hosted a CEO dinner with BCG in Chicago on Thursday where we talked about opportunities in AI, along with the realities of driving change in a more complex threat environment. Here are some highlights from the group discussion and my on-the-record conversation with Admiral Michael Rogers, former head of the National Security Agency and U.S. Cyber Command, who’s now a senior advisor at Brunswick Group. As he put it: “We have to optimize ourselves to deal with a high level of uncertainty, a high level of complexity and a low tolerance for failure by the people that we work for.” Some advice: 

Understand the administration’s agenda. “I encourage everybody to read the 29-page national security strategy this administration wrote,” Rogers said. “They have laid out exactly what they’re doing, and they’ll tell you exactly why. Trump doesn’t care about previous relationships or historic norms. He will tell you, ‘What I care about is what generates value for the citizens of our nation.’ …You cannot be distracted by the rhetoric. You’ve got to focus on what he’s trying to achieve.” 

Understand how the war has changed. Brute military force was the U.S.’s strategy in the past. “Now, we have pivoted to the economic piece. That’s the idea of the blockade, of grabbing Iranian ships around the world…If I was a betting man, I’d say the Americans are going to blink first,” Rogers said. “Whenever this is over, we’re not going back to the way it was …and you can apply what happened in the Strait of Hormuz to the Strait of Malacca, the English Channel, the entrance to The Baltic and other choke points through which massive segments of the global economy pass.”  

Understand how your strategy must change. “You’re seeing cyber attacks at a speed, scale, and dynamic rate of change you just have not seen before,” said Rogers. “You’ve got to spend more time focusing on resilience and less on this idea that it’s all about protection and stopping from getting them in.” Speed and trust were themes echoed by executives around the table. BCG partner Julia Dhar shared some key principles from the research underpinningHow Change Really Works,a new book she co-authored that comes out May 19. (Data on giving people agency especially resonated with me.) And when asked what keeps them up at night, two leaders said they are most worried about the world we are leaving for their children.  

Other takeaways: A lot of the conversation focused on how CEOs are incorporating AI into their operations. At one firm, AI is freeing up staff to have 20% more hands-on time with customers. We talked about how leaders are using AI to boost safety, reduce burnout, and personalize education. And we explored how they’re dealing with broader societal issues of rising fuel costs, food insecurity and distrust. As one CEO said: “We’re making decisions at a faster pace to bring value to our customers … AI is a tool, but your people are going to drive your business.”

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

This story was originally featured on Fortune.com