DeepSeek engineer's essay sparks debate: OpenAI would rather shut down than gate models
soumitrashukla9 · x · 2026-09-15
A retweet chain kicked off a debate over model access philosophy. An essay from a DeepSeek engineer argues OpenAI and Anthropic publicly disagree on a fundamental point: OpenAI works hard to make its most powerful models available within weeks or months of creation, and Sam and Greg would rather shut the organization down than become a SaaS company selling access to a small set of trusted firms — even making negative-EV business decisions to pursue broad, utility-like access.
Peter Gostev adds historical context: before OpenAI, the only models developers could use were things like BERT — if you weren't an ML engineer, you got nothing — while Google, the presumed AI leader, shipped almost nothing usable and kept its progress internal.
Related event: DeepSeek engineer's essay sparks debate over open AI model strategies(2 posts)→
More from Companies & People
- Tailscale called a 'single point of failure' for startups as acquisition chatter splits devs — TheZachMueller · 2026-11-02
- Ex-Shopify engineer: the smartest people I knew all left for frontier AI labs — smlpth · 2026-11-02
- Harvard releases 6 free AI lectures covering GenAI, prompt engineering and RAG — Aiden_Tech_Ai · 2026-11-02
- HUMAIN partners with Microsoft to bring HUMAIN ONE into Microsoft 365 — HUMAIN · 2026-11-02
- Grok, Cursor and X account merge creates absurd login loop for users — oran_ge · 2026-11-02
- Gradio teases NeurIPS 2026 ML Intern paper-reproduction hackathon — Gradio · 2026-11-02