Dev Slams New DeepSeek Model as Distilled Claude Without the Intelligence
Aryvyo · x · 2026-09-19
A developer on X argues the new DeepSeek model feels like a heavily distilled Claude that lost the original's intelligence, calling it at best a "worker model". An individual impression without benchmarks or evidence.
Related event: Developer Slams New DeepSeek Model as Distilled Yet Less Capable(2 posts)→
More from Fun
- Pre-LLM CS Grads Can Never Be True Vibe Coders — They've Seen Too Much — PMinervini · 2026-09-19
- Schmidhuber reignites priority disputes: claims 3 Turing Award winners republished his ideas uncredited — SchmidhuberAI · 2026-09-19
- His most-retweeted post ever is still a 2020 GPT-3 conversation about God — Merzmensch · 2026-09-19
- Fan wires meme AI 'Jev' into an Evangelion MAGI-style three-personality council — i_dg23 · 2026-09-19
- Developer hooks classification-only model Jev to Slay the Spire 2, reaches first boss for 3 cents/hour — Zboubkiller · 2026-09-19
- X user proposes jailing top AI safety expert for 30 days with $1M of OpenAI credits to test escape — wen_ragnarok · 2026-09-19