AI FOMO frenzy, macro update, Fox vs Dominion, US vs China & more with Brad Gerstner

The episode opens with light chatter: a speaker teases someone about running a marathon, someone says they will miss the poker game, and the group runs through a poker-night menu that includes foie gras, a rutabaga salad, duck breast and butterscotch panna cotta. A joke about "austerity" follows, with the market being down and the caviar placed at the bottom of a dish.

The host introduces the lineup: one regular is said to be on Wall Street taking a company public, so the guest Brad joins to discuss macro. The host is on a food tour of Japan and jokes about monetizing a food blog and a TikTok account, then praises Japan's food and how affordable it feels given the exchange rate.

Several speakers compare favorite travel destinations (Italy, Spain, Japan). One recalls taking the older children to Japan about five years earlier and says the Japanese were warm toward large families given a negative birth rate there. Brad says he has been to Japan only twice, when he was poor, staying in a tiny room and eating poor food.

The conversation turns to a Wall Street Journal op-ed that a speaker says DeSantis wrote after Disney's special administrative status was revoked. As described by the speaker, the piece argues that "wokism" is modern Marxism that must be defeated, and that Disney should act like a business answerable to shareholders rather than feed vocal minorities inside the company. Those are the speaker's characterizations of the article, not something checked here.

One speaker argues that DeSantis does well with the Republican base because he accepts that a fully hands-off, Reagan-style approach is not enough when, in this speaker's view, the radical left advances its agenda through legislation, corporations, ESG and private institutions. In this telling, he is willing to use government to push back on behalf of parents, which differs from the Republican approach of 30 or 40 years ago.

Another speaker points to a press tour by Bill Maher, who in an interview with Jake Tapper drew a line between liberalism and wokism, including on trans rights. The host sets that aside, noting the show has a full docket before it reaches the culture wars and the presidential race, and moves to private versus public markets.

The first docket item is what the host calls a massive AI fear-of-missing-out frenzy. An Economist piece cited by the host counts about 500 generative AI startups that have collectively raised more than $11 billion, a figure the host says excludes Microsoft's roughly $10 billion into OpenAI. The host calls the moment a multimedia re-run of the 1990s PC era and a full platform shift.

A clip of Doug Leone of Sequoia is played. He calls AI the next platform shift after mobile and the internet, but warns that investors will overestimate it in the short term and "invest in everything," as they did in 1999. He urges firms to choose investments through more rational market maps rather than because other venture firms are doing everything.

Asked for a reaction, Chamath cites the Munger line about incentives and says a Credit Suisse offer of 6.5% on a three-month T-bill is a sign that the risk-free rate is high. On his reasoning, a long-horizon startup investment must return roughly 20 to 25% to make sense, and limited partners will ask why they pay a 2% management fee if they can earn 6.5% without taking risk.

His conclusion is the opposite of Leone's advice: good investors with a record of distributions may follow it, but most will be under pressure to deploy, so many of the 500 companies will get funded and many will be "torched." He warns of high correlation and no time diversity, meaning too much capital placed in a short window and overpaid for, as he says happened with web 3 and crypto.

Brad, asked how he looks at this across private and public markets, agrees with Leone that this is a platform shift on the scale of the internet or mobile, possibly larger. He puts $10 to $11 billion in context by noting that Meta will spend about $20 billion in one year on AR and VR, and says a technology this large deserves heavy investment even if it is messy. He also agrees with Chamath that the cost of capital limits funding, leaving dry powder on the sidelines.

He draws on the year-2000 analogy: it was clear the internet would be big, but money put into Yahoo, Excite, Lycos and similar names went to zero. Likewise, he says, it is nearly impossible to know which foundation-model company among OpenAI, Anthropic, Cohere, Character, Stability and others will capture value, and the market may settle into an iOS-versus-Android pattern dominated by Microsoft and Google.

On the model layer he says it is hard to pick just one, particularly when the largest one is, in his words, captive to Microsoft, which caps investor upside. He expects big winners in applications and tooling. He adds that he and Chamath have studied the space for a decade or more, and that the transformer model really did change the game.

The host presses Sacks on a contrast: a week earlier he said web 3 and crypto had not stuck for him, yet he quickly listed several compelling AI use cases. Asked whether AI is bigger than mobile, Sacks says he agrees with Brad and Doug that it is the next major platform shift, comparable to mobile, social or cloud. He accepts that nobody yet knows where value will accrue and that it may flow to big companies able to invest heavily.

He adds that incumbents today have talented engineers and will not sit still as they did twenty years ago. Still, he argues, a new platform creates openings for startups. Doug is right that there will be a lot of spray and pray, but Sacks calls it fairly efficient for the ecosystem: any smart engineer with a decent idea gets funded, a Cambrian explosion follows, and most ventures go extinct while some good ones survive.

His own firm, he says, prefers to be selective and put more money behind fewer bets, in line with Doug's tough-love advice. Even so, he sees value in seed funds that plant many flowers and see what blooms.