If OpenAI skips GPT-4.5, this resolves to N/A.
Otherwise starting from the first time it appears at all on the leaderboard, resolves YES if it hits the top spot, and NO if it does not.
Does not apply to any future versions that are also called "GPT 4.5-preview", just the first iteration to appear on the LLMSys leaderboard. If multiple appear at once (like Opus/Sonnet/Haiku), any of them count.
Update 2024-30-12 (PST): - If multiple models are tied for the #1 rank, the top spot will be determined based on ELO rankings. (AI summary of creator comment)
1,000๐ Top traders
| # | Name | Total profit |
|---|---|---|
| 1 | แน746 | |
| 2 | แน57 | |
| 3 | แน52 | |
| 4 | แน49 | |
| 5 | แน35 |
@Bldrt maybe but take a second to notice, the scaling laws are dead, done, finished. Even though the arrow always pointed up. Even though all the smart people in AI labs thought they'll continue to AGI. The scaling laws are dead.
Now, that doesn't mean we won't reach AGI due to algorithmic innovations or reasoning or anything else. but it's worth noticing when you believe in something and it dies.
Sounds good