Author Argues Next-Word Prediction Reached Superhuman Level in 2019
The author argues that LLMs reached superhuman performance on easy next-word prediction by 2019, leaving only hard predictions like math answers and plot twists, which require genuine world understanding that interpretability research confirms exists inside models.
2026-10-19 ~ 2026-10-19 · 2 related posts
- LLMs have been superhuman at next-word prediction since 2019 — only hard words remain — mlpowered · 2026-10-19
- Only hard words left: predicting twists and conclusions forces models to build world understanding — mlpowered · 2026-10-19