> ## Content Index
> Fetch the complete content index at: https://wire.fourthweb.ai/llms.txt
> Use this file to discover other available public pages before exploring further.

# Billionaires Cash Out $3.3B While Retail Bets on AI
- URL: https://wire.fourthweb.ai/billionaires-cash-out-3-3b-while-retail-bets-on-ai/
- Published: 2026-10-16T19:02:09.000Z
- Updated: 2026-10-16T19:02:12.000Z
- Description: Three billionaires just hit the exit button while the AI true believers are still buying the dip. CoreWeave's three co-founders have sold $3.3 billion in shares since the company's IPO, timing their exit amid rising credit risks and investor skepticism about AI infrastructure spending
- Author: Travis Wright
- Tags: AI Agent Economy, AI Agents, AI Infrastructure, Tokenized Assets, IPO Watch, Big Tech

**Three billionaires just hit the exit button while the AI true believers are still buying the dip.**

### The Summary

- [CoreWeave's three co-founders have sold $3.3 billion in shares since the company's IPO](https://www.bloomberg.com/news/videos/2026-10-16/only-third-inning-relative-to-this-ai-game-says-ives-video?ref=wire.fourthweb.ai), timing their exit amid rising credit risks and investor skepticism about AI infrastructure spending
- Dan Ives maintains we're only in the "third inning" of AI adoption, even as insider selling signals less confidence than the bulls want to admit
- The gap between what analysts say publicly and what founders do privately is worth watching

### The Signal

CoreWeave went public betting that the AI infrastructure gold rush would never end. The company became the poster child for [GPU](https://wire.fourthweb.ai/tag/compute-wars/)\-as-a-service, renting [compute](https://wire.fourthweb.ai/tag/ai-infrastructure/) power to companies training foundation models. [Now three co-founders have collectively dumped $3.3 billion in stock](https://www.bloomberg.com/news/videos/2026-10-16/only-third-inning-relative-to-this-ai-game-says-ives-video?ref=wire.fourthweb.ai), the kind of insider selling that doesn't show up in press releases but speaks louder than any earnings call.

The timing matters. CoreWeave's stock has been volatile as credit markets tighten and investors start asking harder questions about AI capital expenditure. When does training the next frontier model stop being a necessity and start being a vanity project? When do companies realize they're paying premium prices for compute they don't actually need?

> "We're only in the third inning relative to this AI game." — Dan Ives

Ives is still bullish, framing current volatility as noise in a longer transformation. He's not wrong that we're early in AI adoption. Most companies haven't deployed agents at scale. Most industries haven't figured out how to tokenize their operations or build AI-native workflows. But "early" doesn't mean "guaranteed upside for everyone who showed up."

The real question is what kind of early we're in:

- Early like Amazon in 1998, where infrastructure players compound for decades?
- Early like pets.com in 1999, where the category is real but most companies are mispriced?
- Early like solar in 2007, where the tech works but the business models don't yet?

CoreWeave's founder exits suggest they're betting on door number two or three. You don't sell $3.3 billion if you think the next ten years look better than the last two. You sell when you think your stock price has gotten ahead of your revenue trajectory, or when you see the capital markets getting skeptical before the rest of the market does.

The credit risk angle is underreported. CoreWeave has borrowed heavily to build out GPU clusters. That works beautifully when interest rates are low and AI spending is unlimited. It works less beautifully when rates stay elevated and CFOs start asking their AI teams to show ROI in quarters, not years. If demand for training compute plateaus or shifts toward inference, companies like CoreWeave are left holding expensive hardware financed with expensive debt.

### The Implication

Watch what insiders do, not what analysts say. If you're building in the agent economy, this is a reminder that infrastructure players face different risks than application layer companies. The people renting you GPUs might be overleveraged. The people selling you AI optimism might be selling their stock.

The AI build-out is real. The hype cycle around who profits from it is also real. Those are not contradictory statements.

### Sources

[Bloomberg Tech](https://www.bloomberg.com/news/videos/2026-10-16/only-third-inning-relative-to-this-ai-game-says-ives-video?ref=wire.fourthweb.ai)