RT @linexjlin: Karpathy 的最后一段说的很好。 他认为现在我们在卷 AI 训练集群规模，数据质量，各种局部调优...，本质上是在追求对人类数据的拟合精度，很可能已经陷入到另外的 “The Bitter Lesson” 当中了。这可能也是 Sutto…
- 0 replies
- 0 recasts
- 0 reactions
RT @LuozhuZhang: > “He sent out an email on Friday evening saying everything is going to deep learning, and that we were no longer a graphi…
- 0 replies
- 0 recasts
- 0 reactions
codex-cli gpt-5-high planning 和 debug 真是强太多，手里的 cc max 突然不香了
- 0 replies
- 0 recasts
- 0 reactions