DeepSeek team told a VC two years ago that media generation was 'unserious'
michellechen · x · 2026-09-22
Michelle Chen recalls meeting the DeepSeek team in China two years ago; asked whether they'd train other models, they said they only want to push the frontier of research and consider media generation unserious entertainment. She notes media gen is itself a gigantic billion-dollar market (e.g., Disney), but research camps clearly split on this, and suspects Anthropic has a similar mentality given its LLM focus.
More from AGI Musings
- Rogue AI is a scapegoat narrative: the HF incident was human error — whurley · 2026-09-22
- Ex-Slack engineer launches 'After Coding' podcast on life when LLMs do the coding — josh_wills · 2026-09-22
- Report: OpenAI has automated much of its experimental model training pipeline — Dr_Singularity · 2026-09-22
- Report: OpenAI has automated much of experimental model training—AI writes GPU kernels, agents collaborate — Dr_Singularity · 2026-09-22
- Tech is just $200k engineers automating each other's jobs while claiming to build AGI — tech__unicorn · 2026-09-22
- Venkatesh Rao essay: EAs are great at numerical uncertainty, bad at deeper doubt — binarybits · 2026-09-22