@MarcusAbramovitch and Remmelt have made a $5k:$25k bet on exactly this question. Links to EA Forum and LessWrong where they announced the bet.
An AI market crash will be defined, for the purposes of this bet, as at least 2 out of 3 of the following criteria being met.
Here are our criteria:
OpenAI's 2025 or 2026 annual revenue is below $1.6 billion.
Anthropic's 2025 or 2026 annual revenue is below $400 million.
Nvidia’s data center revenue in any quarter from now to Q4 2026 is below $8.5 billion.
I'm opening this question purely out of personal curiosity and it has nothing to do with my current employer. I'll resolve the question based on how Marcus and Remmelt decide, and I won't be trading on it since that would clearly constitute insider trading.
1,000🏅 Top traders
| # | Trader | Total profit |
|---|---|---|
| 1 | Ṁ793 | |
| 2 | Ṁ307 | |
| 3 | Ṁ159 | |
| 4 | Ṁ130 | |
| 5 | Ṁ43 |
People are also trading
I'm pretty sure it's literally impossible for this to resolve YES at this point. Based on ARRs, OpenAI and Anthropic both easily hit the 2025 criteria, and in fact both have presumably already hit the 2026 revenue criteria! Even if a solar storm knocked out all electronics for the rest of the year I think this would resolve NO because OpenAI and Anthropic have already hit the revenue criteria for both 2025 and 2026, and Nvidia alone missing its quarterly criteria isn't enough for this market to resolve YES.
@PlasmaPower Agreed. @palcu you can resolve this NO now. Research with Claude confirms this and gives me these estimates:
- Anthropic: Jan + Feb actual revenue ≈ $2.75B (beats $0.4B)
- OpenAI: Jan + Feb actual revenue ≈ $3.5–4B (beats $1.6B)
This is calculated from published annualized revenue figures at various dates. If all three companies went bust tomorrow, the market would still resolve NO.
> I'll resolve the question based on how Marcus and Remmelt decide
Marcus has decided: https://www.lesswrong.com/posts/BYTA3d8SudnpcFery/our-bet-on-whether-the-ai-market-will-crash?commentId=B6MMpr64TCCFv4uWK
But at this point this is math, not a human judgment call.
@Doodle Jump I’d be curious to hear what specific indicators you're using to forecast such a revenue drop, especially for Nvidia given its broader exposure beyond just AI.
Ah, I just figured out that the dates are wrong. FYI traders that this should be "by the end of 2026" and the resolving date will be Jan 1 2027. I've reached out to some moderators to change it.
Nevermind, the date is good, I need another coffee.