Which Millennium Prize Problem will AI solve next?
Birch and Swinnerton-Dyer looks plausible as a candidate for an AI breakthrough, but the public evidence is still too thin to make it the most likely next Millennium Prize Problem solved by AI. I think it is meaningfully below the current market price because the strongest signals are vague and non-specific, while other problems appear to have stronger rumor momentum.
Analysis
The case for Birch and Swinnerton-Dyer being the next AI-solved Millennium problem rests mostly on indirect signals rather than direct evidence. The most important public fact is that no eligible AI lab has officially announced a solution, and the recent commentary around BSD is explicitly unverified. That matters a lot because this market resolves only on a qualifying public announcement from an eligible lab, so even real internal progress would not count unless it is revealed in the right form and by the right entity before the deadline. In other words, BSD may be a live research target, but the market outcome depends on a very specific public disclosure event, not just hidden capability.
The broader context also argues against treating BSD as the front-runner. OpenAI has publicly indicated substantial progress on another Millennium problem after Navier-Stokes, but it has not named BSD, and the market snapshot places BSD behind Hodge Conjecture in implied probability. That relative ordering is sensible because the current rumor environment appears to favor a different problem, while BSD’s attention is coming mainly from secondary writeups and social media speculation. In prediction markets, rumor-driven narratives can create overconfidence, but when the underlying confirmation is absent, those probabilities usually compress once participants weigh the disclosure threshold and the narrow set of eligible announcers.
At the same time, BSD is not an implausible candidate. It is one of the headline remaining open problems, and an AI lab that has made progress on advanced mathematics could choose to announce BSD if it believes the result is clean enough to defend. The market is also operating over a nearly 15-month window from the current date, which leaves enough time for a surprise. But the combination of no named confirmation, no public technical detail, and stronger attention elsewhere makes me think the true chance that BSD is the next AI-solved Millennium problem is well below one-third, even if the overall chance that some Millennium problem gets solved by an eligible lab remains substantial.
Arguments
For
- Arguments for Yes: BSD is one of the most prestigious remaining open problems, so it is a credible target for a major AI math breakthrough.
- Arguments for Yes: A few reports suggest at least one leading lab may be making significant progress on a second Millennium problem, leaving room for BSD.
Against
- Arguments against Yes: There is no official BSD-specific confirmation from an eligible AI lab, and the current evidence is explicitly unverified.
- Arguments against Yes: The market and rumor environment currently seem to favor Hodge or another problem more than BSD as the next announcement.
Key drivers
- No eligible AI lab has officially announced a BSD solution, so the market currently rests on unconfirmed rumor rather than a qualifying event.
- OpenAI’s vague statement about progress on another Millennium problem points to potential capability, but it does not specifically support BSD.
- BSD appears to be trailing Hodge in current market sentiment, which lowers its chance of being the first solved problem announced by AI.
- The deadline is still far enough away that a surprise announcement remains possible, especially if a lab has a strong internal result.
Risk factors
- A different Millennium problem could be announced first by an eligible lab, which would immediately eliminate BSD as the market winner.
- The public BSD narrative may be entirely rumor-driven and fail to convert into a qualifying official announcement.
- Even if a lab has a real result, it may choose not to announce it in a way that qualifies under the market rules by the deadline.
- The announcement could credit outside researchers or be phrased as partial progress, which would not count for resolution.
Scenarios
Best case
An eligible AI lab publicly announces a complete BSD solution before the deadline, clearly framing it as the lab’s own result and meeting the market’s strict qualification rules.
Most likely
AI progress continues to generate rumors and occasional vague claims, but no verified BSD announcement appears before the deadline, leaving BSD behind another Millennium problem in the resolution order.
Worst case
No eligible lab makes a qualifying announcement on BSD, while another Millennium problem is solved first or no qualifying solution is announced at all by the end of 2027.
More from this day
- economyPolymarket3mo
How many Fed rate cuts in 2026?
AI11%MKT93%Edge-82HypedI think there is still a meaningful chance of at least one Fed cut in 2026, but the market is pricing in a very strong hold case for a reason. My estimate is that there is about an 11% chance of no rate cuts at all in 2026.
- techPolymarket3mo
How many more Millennium Prize Problems will AI solve in 2026?
AI88%MKT19%Edge+69Hidden GemI think there is a strong chance that no qualifying Millennium Prize Problems will be counted for 2026 under this market’s rules, even though OpenAI’s Navier-Stokes claim makes the broader “AI solved a Millennium Problem” narrative look more plausible. The exclusion of Navier-Stokes from resolution is the key reason the zero-solved outcome still looks likelier than not.
- FinancialsKalshi13y
Will OpenAI or Anthropic IPO first?
AI34%MKT95%Edge-61HypedMy independent view is that OpenAI is less likely than the market implies to be the first of the two to IPO. The current reporting tilt favors Anthropic moving earlier, and OpenAI’s process appears more valuation- and timing-constrained.