I've been defaulting to asking for most things in Markdown since the GPT-4 days, when the 8,192 token limit meant that Markdown's token-efficiency over HTML was extremely worthwhile.
早期由于token限制，Markdown因其高效性成为首选，但随着模型能力提升，HTML的优势逐渐显现。
3 Matching Annotations
- May 2026
-
simonwillison.net simonwillison.net
-
-
I've been defaulting to asking for most things in Markdown since the GPT-4 days, when the 8,192 token limit meant that Markdown's token-efficiency over HTML was extremely worthwhile.
过去由于token限制，Markdown因其效率优势而成为首选，但现在这一限制已不再适用。
-
- Apr 2026
-
-
Muse Spark compresses its reasoning to solve problems using significantly fewer tokens. After compressing, the model again extends its solutions to achieve stronger performance.
这种思维压缩-扩展的循环过程暗示了AI可能发展出类似人类的抽象思维能力，先提炼核心再展开细节，这一发现对理解AI推理机制和未来优化方向具有重要启示。
-