Researchers slam frontier LLM writing quality: "getting worse with every version"
arjunrajlab · x · 2026-10-17
Replying to Stanford professor Anshul Kundaje's rant — "can someone please train a frontier LLM that is actually good at LANGUAGE... they seem to be getting worse with every version" — researcher arjunrajlab says Gemini 3 Pro was a high-water mark for writing, and that Astra might be the current best, promising to try it. Reflects academic users' frustration with declining prose quality in frontier models.
More from Models
- AI scientist slams Anthropic: Claude anthropomorphization is dangerous, secrecy betrays science — rbhar90 · 2026-10-17
- Hash Rate Ep. 189: Jev models on Bittensor subnet SN92, decentralized training vs traditional LLMs — markjeffrey · 2026-10-17
- 'Remove AI tics and mannered prose': a one-line prompt fix for Argon's writing — arjunrajlab · 2026-10-17
- Blind test of 200 real image edits: nano banana pro loses to its cheaper base model — ravisparikh · 2026-10-17
- Dev ditches Gemini 4 Argon for Grok 4.5, which eats 96.6% of his Cursor tokens — gaganghotra_ · 2026-10-17
- LMArena opens Code Arena WebDev for blind head-to-head model coding battles — arena · 2026-10-17