Andrej Karpathy 12/10/2025

Auto-grading decade-old Hacker News discussions with hindsight

Read Original

The article details a project that uses LLMs (like ChatGPT 5.1 and Opus 4.5) to automatically analyze Hacker News front pages and comment threads from 10 years ago. It evaluates the prescience of past discussions, awards 'most prescient' and 'most wrong' comments, and reflects on how future AI might scrutinize all digital history. The project's code is shared on GitHub.

Auto-grading decade-old Hacker News discussions with hindsight

Comments

No comments yet

Be the first to share your thoughts!

Browser Extension

Get instant access to AllDevBlogs from your browser

Top of the Week