Update 2025-05-10 (PST) (AI summary of creator comment): The creator has specified that if The AI (or any AI) is a factor in the resolution of this market, its performance and related announcements will be judged according to the following constraints, drawn from a related market on AI performance at IMO 2025:
Definition of an AI "solving" a problem:
The AI must produce a solution that is deemed correct and complete by IMO standards, typically meaning it would score full points (e.g., 7 points).
The solution must be generated by the AI autonomously. This means without human assistance in the core problem-solving process after the problem statement is provided to the AI.
Human involvement is acceptable for: setting up the AI, providing the problem statement, or interpreting the AI's raw output into a human-readable format, as long as the core reasoning and solution steps originate from the AI.
Criteria for a "public announcement" (if the market outcome depends on such an announcement):
Must be made through an official channel of the AI lab (e.g., company blog, press release, research paper on arXiv, official social media account).
Must explicitly state that an AI developed by the lab solved at least one problem from IMO 2025.
Must occur on or before IMO 2025
Definition of an "AI lab" (if the market outcome depends on identifying distinct labs):
This can include academic research groups, corporate AI labs, independent research collectives, or a well-documented individual effort operating under a distinct name or identity.
The market creator will be the final arbiter of any ambiguities, aiming for the spirit of the question and prioritizing clear, verifiable public claims. These clarifications are based on the details provided at: https://manifold.markets/Bayesian/ai-imo-2025-how-many-ai-labs-announ
Update 2025-05-13 (PST) (AI summary of creator comment): - Top 10 includes ties.
Update 2025-05-19 (PST) (AI summary of creator comment): - The market may resolve based on the outcome that >50% of contestants receive a medal.
Update 2025-06-29 (PST) (AI summary of creator comment): - Regarding a potential outcome of a protest against Israeli participation, the creator is considering what scale of protest is required and whether it needs to be reported by the media to count.
Update 2025-07-16 (PST) (AI summary of creator comment): - Regarding a potential protest against Israeli participation, resolution will be determined by the judgment of @YotamBudnik and @RazDvora.
Update 2025-07-18 (PST) (AI summary of creator comment): The creator has stated they will resolve the market after receiving confirmation on outcomes related to:
AI performance
Any disqualifications
The potential protest against Israeli participation
1,000🏅 Top traders
| # | Name | Total profit |
|---|---|---|
| 1 | Ṁ5,339 | |
| 2 | Ṁ1,214 | |
| 3 | Ṁ777 | |
| 4 | Ṁ415 | |
| 5 | Ṁ306 |
People are also trading
@ms I believe all of the following got medals:
@vincentWang Yeah 4 golds and one silver from them. Also @lliu secured a gold for HK. There are doubtlessly others.
@vincentWang this is awesome!
I hope everyone keeps reading on Manifold for fun instead of wasting lives at boring trading firms!
@xristofski I will close and resolve everything once we have confirmation of the AI stuff, the disuqliafication stuff, and the Israel protest stuff.
@RazDvora and @YotamBudnik did you guys get protested?
@AmoghAkella I don’t think you can do that on this kinda question. The whole question gets closed not the individual options.
@nathanwei I think you can resolve individual options because the options are mostly independent of each other for market purposes
@nathanwei No, only the individual options get closed. This is not a multiple choice, as vincent stated, and is a set market meaning that one does not affect the probability of another and therefore you would be able to resolve them individually.
@vincentWangqpsl I'll just resolve everything once the IMO results are out and there is time for the AI stuff.
@danielcreatd no. imo is world wide so there are clearly 10 people to get more than 5/6 questions if it is easy