Gemini models were announced on December 6th. Gemini Ultra slightly outperforms GPT-4 across the suite of benchmarks, with some caveats (ex. CoT@32 vs. CoT@8).
Was the Gemini Ultra choice and announcement at this state intentional to slow the capabilities race? This questions resolves to "Yes" if evidence surfaces that the Gemini Ultra model was chosen to be announced in order to not hasten the capabilities race. If more capable models were produced but not announced instead, this does not necessarily resolve to "Yes", only if evidence from within Google confirms that other models were not released in order to not hasten the capabilities race.
Edit: Resolution date extended to 12/31/2024!
1,000
3.00@horace It varies but generally market makers have a lot of leeway to update markets to better fit their original intent. Since there is only 157 Mana in the market and you didn't explicitly say when the evidence has to arrive by, I don't suspect anyone will have an issue if you extended the resolution date.