DeepSeek kernel engineer's emotional blogpost on open source vs AI doom
teortaxesTex · x · 2026-09-17
Shengyu Liu (interestingLSY/intlsy), a kernel engineer at DeepSeek, published an emotional Chinese blogpost: "Because AI may destroy human civilization, I choose to believe in a faith that can truly benefit humanity, and stay at DeepSeek."
The first part describes his personal struggle with his work rendering his beloved craft obsolete; the second lays out his ideals. The retweeter uses it to argue against the AI doom narrative: knowledge should be accessible and cheap — "if AI is dangerous, more people should understand it, control it, and balance that power together," something figures like Dario may never grasp about why open source matters.
The post is being read as a rare personal statement of DeepSeek's open-source culture.
Related event: DeepSeek Kernel Engineer's Essay on AI Risk and Staying On(2 posts)→
More from AGI Musings
- Long-read: AI's real value comes after the efficiency era—infra glut, agent tooling boom — yangyi · 2026-11-02
- Security's attack-defense automation gap mirrors math: proofs are easy, conjectures are hard — mboehme_ · 2026-11-02
- vikhyat predicts alternating compute oversupply and undersupply on the road to AGI — TheZachMueller · 2026-11-02
- Sam Altman: AI has surpassed top scientists, and 2026 will be a turning point year — kimmonismus · 2026-11-02
- DeepSeek design doc: a person can only follow 3-5 agent sessions at once, attention is the real bottleneck — heyshrutimishra · 2026-11-02
- AI researcher: we have no reliable way to tell whether an AI system is conscious — dioscuri · 2026-11-02