LLMs have been superhuman at next-word prediction since 2019 — only hard words remain
mlpowered · x · 2026-10-19
mlpowered argues that when people imagine next-word training they picture phone autocomplete-level tasks. But LLMs have been superhuman at next-word prediction since 2019 — they long ago ran out of easy words. What remains are hard ones: the answer to a complex math problem, a novel's twist, a PhD thesis's conclusion — predicting those well requires building world understanding.
Related event: Author Argues Next-Word Prediction Reached Superhuman Level in 2019(2 posts)→
More from Models
- NVIDIA open-sources structured-data-models: foundation models for tabular and time series data — lmoroney · 2026-10-19
- DeepSeek V4.1 Flash becomes first open model to crack top 10 on search leaderboard — gaganghotra_ · 2026-10-19
- KV Cache Compression Makes 1M Context Manageable at Under 100GB Per User — khademinori · 2026-10-19
- VoxelBench 2 nears beta as Fable 5.5 posts a monster run on voxel builds — koltregaskes · 2026-10-19
- Claude Opus 5.5 lists what it wishes the training corpus had contained — repligate · 2026-10-19
- Grok Sycophancy Scandal: Picks Musk Over 5M Slovaks, Hails His '100-Hour Week' Boxing Endurance — john__allard · 2026-10-19