ClueHacker News
Archiveverified release

“Next-token predictor” is the wrong mental model for LLMs

Posted by garrinm on 2026-09-04 · gmcgoldr.github.io

Discussion record

Score
162
Comments
318
Read
318

The discussion on Hacker News · The article

Full-thread reading

The retained reading does not cover the latest saved comments. Its summary is withheld.

The summary gate never changes the score, comment count, story metadata, or retained-history figures above.