Name: Is The AI Bubble About To Pop? - Chamath Palihapitiya
Uploaded: 2025-08-24T15:00:15+00:00
Duration: 577 s

Speakers

David Sacks — General Partner, Craft Ventures

Chamath Palihapitiya — CEO, Social Capital

Jason Calacanis — Angel Investor / Founder, LAUNCH

Summary

The All-In hosts debate whether the AI trade is breaking after a week in which AI stocks sold off on a viral MIT study, Sam Altman's bubble comments and Meta's AI hiring freeze. Chamath argues the first wave of enterprise AI spending was undisciplined board-driven budget deployment, that back-office automation is the high-ROI use case while sales and marketing tools fail, and that a sorting and cleansing of AI application companies is coming just as it did in social and SaaS. David Sacks counters that the roughly 10% correction in public AI stocks was healthy rather than the start of a bust, and that the AI investment super cycle is still intact. He adds that model performance is clustering instead of racing toward recursive self-improvement, so AI should be treated as a normal technology race by investors and regulators alike.

MIT study: 95% of generative AI pilots never reach production.
Chamath: first-wave AI budgets were board-driven and poorly allocated.
Back-office automation shows the highest AI ROI; sales and marketing lag.
Chamath expects churn and consolidation in AI apps, as in social and SaaS.
Sacks: the ~10% drop in public AI stocks was a healthy correction, not a bust.
Sacks: the AI investment super cycle and boom remain intact.
Model performance is clustering; GPT-5 fell short of lofty AGI expectations.
Overhyped AGI narratives triggered a policy backlash and AI safety bills.

Ideas

Chamath Palihapitiya CEO, Social Capital 2:26

AI app boom faces shakeout and churn.

Chamath says the first wave of enterprise AI spending was boards reading the word AI, asking CEOs for an AI strategy, and large existing budgets being sprayed across experiments, and that the industry is now entering the sorting function. He questions how durable the resulting revenue is: companies reaching $50 million of ARR within months and raising huge rounds have not yet been tested for logo churn or dollar churn, and that revenue can be taken back by cheaper new entrants, by the foundation models moving up the stack and absorbing the capability, or by projects that simply do not work and get abandoned. He points to precedent: seven or eight thousand social companies collapsed to about five winners within six years, and SaaS took many years to sort out a handful of winners after early successes like Yammer. He suspects AI is about to go through the same sorting and cleansing before the sector rebuilds from first principles, which makes the AI complex a regime to monitor rather than a clean buy here.

David Sacks General Partner, Craft Ventures 4:29

AI selloff healthy; super cycle intact.

Sacks argues the roughly 10% pullback in public AI stocks was a healthy correction in sentiment rather than the start of a bust cycle. In his view the selloff deflated the fantastical claims that AGI was two to three years away and that recursive self-improvement would let one model capture all the value, while leaving the underlying investment super cycle intact: he still sees a boom, still sees a lot of interesting AI applications being developed, and still believes AI is a new and important form of computing that will unlock tremendous value in the economy. The qualifier is timing, not direction: making AI generate business value takes prompting, iteration and validation, so the payoff arrives over a longer and more normal adoption timeline than the hype implied. On that basis he treats the MIT pilot-failure study and the mixed GPT-5 reception as a healthy dose of skepticism applied to overhyped claims, not as evidence that the AI trade is over.

Is The AI Bubble About To Pop? - Chamath Palihapitiya

Summary

Ideas

Up Next