Vague DeepMind model teasers come from PMs, not the engineers actually training models
nagpalchirag · x · 2026-08-22
The author observes that vague posting about model progress at Google DeepMind usually originates from Program Managers, who are likely quite detached from actual model training, while engineering staff stay silent. This either reflects a PM-led Google culture, or the engineers know the brass tacks and won't put their public reputation on the line. Their suggestion: having actual engineering staff post about model development would be better marketing for Google.
More from Companies & People
- AI Whisperers launches online robotics course, citing 10x complexity increase — StewartalsopIII · 2026-08-22
- Mystery model Ox Alpha sparks debate: Could reshape US-China AI landscape — brandon_galang · 2026-08-22
- Theory: OpenAI's Astra to be first end-to-end trained multi-agent orchestrator — daniel_mac8 · 2026-08-22
- Anthropic hires ex-TPU leader Amir Salek to develop custom AI chips — Pablomiller · 2026-08-22
- Report: Musk struck deal with ASML to purchase massive equipment — burhop · 2026-08-22
- Dev Maintains X App Despite Toxic Environment and Criticism — Dimillian · 2026-08-22