Will OpenAI's next-generation model score 65% or higher on the GPQA benchmark?
Basic
12
303
2027
67%
chance

Resolve to YES if OpenAI's next generation language model scores 65% or higher on the GPQA benchmark(extended set).

If OpenAI's existing model gets 65% or higher by post-training enhancements, that also counts.

There's room for improvement via prompt engineering after the release, but I don't know how long I should wait, so I will resolve this question as soon as OpenAI releases their next model.

Get Ṁ1,000 play money