Google currently leads with Gemini -- which has two models at around 1370
But OpenAI just announced O3 -- which is getting great marks on things like hard science questions.
https://deepnewz.com/ai-modeling/openai-unveils-o3-o3-mini-models-exceeding-human-performance-on-arc-agi-4f05e4f7
The resolution is simple. Will and LMSys update contain a model with 1400 ELO? Cutoff is last day in January (East Coast time).
Update 2025-26-01 (PST): - Resolution Criteria Update:
The resolution will be based on the information available on the website on February 1st. (AI summary of creator comment)
@Moscow25 is it per status on the last of January, or will you include whatever's the first update in February? I think you usually go by the latter on your markets (am I conflating creators, perhaps?), but unless it's in description I imagine it should be by the former.
You can bet on which one crosses 1400 first here
https://manifold.markets/ChinmayTheMathGuy/what-will-be-true-of-the-first-mode
Worth noting: this market is essentially https://manifold.markets/bobbill/will-any-llm-outrank-gpt4-by-150-el but with a 1 month later close date