Noam Brown· @polynoamial · X·· 2026-04-24AI 评分68
AI 导读
我至死坚持的观点：对于当今的AI模型，智能是推理计算的函数。自2024年起，用单一数字比较模型已不再合理。重要的是每个token或每美元的智能。 在像Codex这样的产品中使用时尤其如此。
正文 · 原文
A hill that I will die on: with today's AI models, intelligence is a function of inference compute. Comparing models by a single number hasn't made sense since 2024. What matters is intelligence per token or per $.
This is especially true when using it in a product like Codex.
The GPT-5.5 model family completely dominates the cost-performance frontier on the Artificial Analysis Index在 X 查看被引用的帖子
来源：Noam Brown · x.com