Most Replayed Moment: AI Is Changing How Teams Work Forever

The host begins by asking Nick Hanauer for the background behind his unusual perspective as a billionaire entrepreneur. Hanauer says he was raised in a civic-minded, middle-class family. His father joined a small family business founded by Hanauer’s grandfather that manufactured bed pillows and down comforters, giving Hanauer an early connection to business. He later developed an intuition that the internet would transform commerce. A friend, Jeff Bezos, shared that intuition while working in New York for a hedge fund, and Bezos eventually sent his belongings to Hanauer’s house in Seattle. The two then started Amazon.com together. Hanauer describes the origin story as involving both conviction and luck. Daniel Priestley says he grew up in Australia and discovered entrepreneurship as a teenager after finding an entrepreneurial mentor. He spent two years in a startup, which made entrepreneurship feel like a “cheat code” for life. At 21, he started his first agency; it grew from zero to $1 million in its first year and $10 million in its third year. He recalls it as an unusually immersive young company, with employees living and spending much of their time together. Priestley says he subsequently became interested in entrepreneurship, small businesses, and family businesses, and started an entrepreneur accelerator 15 years ago to help people learn how to run companies.

The host frames the central question with recent warnings from Anthropic and a graph showing declining entry-level job postings. He asks how disruptive AI will be and whether it will cause widespread job losses. Hanauer says Bernie Sanders’s proposal that the United States own 50% of AI companies is not necessarily unreasonable. His argument is that AI is monetizing humanity’s intellectual property for free while concentrating the gains among a small number of people. He compares the proposal to Norway’s sovereign wealth fund, suggesting that a fund capturing a significant share of AI-created value could recycle it into society and cushion the disruption. He says this idea could make sense for the United States and China, though not necessarily for countries such as Senegal or Southampton. Priestley agrees that places such as the Philippines, which benefited from outsourced back-office work, could be badly affected as AI performs those tasks. But he offers a more local and practical response: the United Kingdom has about 5.7 million businesses and roughly 1 million unemployed people, so if one-fifth of businesses hired one additional person, the numbers would line up. He suggests helping small businesses use AI for marketing, legal contracts, and other tasks, while offering training and tax incentives for companies that hire. The host then points out that AI agents can operate computers, perform data entry, edit material, and make cold calls—the kinds of tasks often assigned to entry-level workers. The question becomes what entry-level employees will do if businesses become able to automate those duties.

Priestley says his own group of small, dynamic businesses has implemented AI and hired people as a result. The new employees are sometimes entry-level workers whose productivity is augmented by AI. AI-generated appointments have led him to hire more salespeople, while AI-assisted marketing creates more opportunities for people who can conduct the final human conversations. He describes an “AI layer” inside his organizations that combines context, skills, models, and security, then produces information about whom the company should contact, why they should contact them, and what they should say. Hanauer accepts that AI could make one person capable of doing the work of five, but says a company has a choice: eliminate four jobs or retain the employee, equip them with better tools, and outperform competitors. He compares the current anxiety with the introduction of calculators and computers. People once worried that calculators would make accountants or mathematics education obsolete, yet computers ultimately increased the amount of work people performed rather than simply reducing it. In his optimistic view, doctors will become better doctors and other professionals will be augmented. Priestley pushes back on the comparison because AI is being distributed instantly at global scale. When Anthropic releases a new model, it can reach users everywhere at once, unlike the slow and expensive rollout of household computers. He says his team does not intend to dismiss anyone because of AI and is instead reskilling employees, but admits that companies may hire fewer people in the near term. He cites a call center that is allowing a natural annual attrition rate of 25% to reduce its workforce without replacing departing staff, and says Klarna’s CEO described using the same approach.

Priestley offers a counterexample to the attrition story. He works with a husband-and-wife team in northern England who operated a small video-production agency with only one or two contractors. They used AI to build software that automated scriptwriting and related production tasks. After launching a waiting list, they attracted 5,500 people. They then acquired their first 1,500 customers for software that had cost almost nothing to build in only four months, and they are now hiring a team of 10. Priestley argues that this kind of growth would previously have been difficult for such a small business. The couple did not need to raise millions of dollars or hire 30 or 40 people before testing the opportunity. AI made it possible for them to turn a constrained service business into a rapidly growing software company. Hanauer agrees that he could produce many anecdotes of entrepreneurial opportunities created by AI, but says anecdotes do not resolve the larger economic question. The speakers distinguish between individual opportunity and economy-wide disruption. Priestley says the nature of value creation is changing quickly and that many people may have to transition into new types of work. He observes that his own recruitment process increasingly looks for people with particular AI-related skills, yet those candidates are becoming harder to find. He and the host agree that education and training are necessary, although Priestley emphasizes that training takes time while the technology is moving immediately.

Hanauer compares the coming transition with the Jevons Paradox: a tractor could displace 100 agricultural workers, sending them into cities to look for work at once. He links that earlier disruption to the social turmoil depicted by Charles Dickens and to Karl Marx’s emergence as a critic of the resulting economic system. Asked what should be done now, Hanauer rejects universal basic income for the moment, but argues that society must help people manage the transition. The valuation of AI companies, he says, depends on the expectation that AI will disrupt and replace jobs. If that value is created by displacement, some of it should be captured and recycled into the economy. The host questions whether this is socialism. Hanauer says it is closer to common sense and argues that high-functioning democracies already manage markets for public purposes through progressive taxation and labor standards. He distinguishes between seizing private property and owning strategic assets. The host raises Dubai, where the government owns major physical hotel assets and leases them to operators. Hanauer replies that Dubai developed those assets, whereas AI companies may have built their systems from data that originated with the public. Hanauer suggests treating data as a common asset—“the new oil”—and charging companies a license for using it. He says these companies drew value from people in Africa, Britain, Australia, and Canada. The host asks whether that would mean taking a portion of AI companies, and Hanauer says a sovereign wealth fund could be one way to capture the value without abandoning markets. He remains skeptical of explicitly socialist proposals, while Hanauer says Bernie Sanders is not proposing to abolish markets but to manage them for public benefit rather than only for capital owners.

The discussion turns to whether the same public-benefit logic should apply to Amazon and other large companies. Hanauer says governments already take a portion through taxes, but argues that the current taxes are insufficient. He says Amazon has been highly successful at avoiding taxes and supports closing loopholes so that it pays taxes like other businesses, including the small retailers competing with it. The host asks whether the deeper problem is that governments are not capable of managing this transition. Priestley agrees, citing a statistic about the UK government: an employee is reportedly 10 times more likely to die than to be fired for poor performance, and the government fires people for incompetence at one-six-hundredth the rate of ordinary businesses. Priestley points to Singapore as a model because its government emphasizes meritocracy, promotion, and dismissal based on outcomes. Hanauer calls Singapore a “miracle of governance,” while acknowledging that it is a small place and perhaps the only country to benefit from a well-meaning dictator. They agree that the AI transition will produce both job disruption and new kinds of work, but that education and training must catch up. Priestley says he now asks job candidates how deep they are in the “AI rabbit hole”; someone who says “a lot” immediately signals strong fit for his team. The final examples widen the forecast beyond software. Priestley mentions humanoid robots, including a Figure AI video showing a robot sorting packages on a production line for eight straight days, as well as autonomous vehicles, including his self-driving car in Los Angeles and driverless taxis. He expects the future to feature small teams—perhaps 10 people making a YouTube channel or software product—rather than only huge corporations. Despite their disagreement over policy, the speakers converge on the idea that knowing which work to automate and which decisions still require people will be crucial.