OpenAI co-founder Trask: AI framing bakes in power concentration; fix it with federated infra
iamtrask · x · 2026-09-02
OpenAI co-founder Andrew Trask discusses power concentration with scholar Seth Lazar. Lazar argues concentration isn't inherently bad, but research shows it raises the odds of unjustified use of power, so it should be avoided where possible. Trask responds that unnecessary concentrations of power are net-bad, and the real debate is when concentration becomes unnecessary. He argues the framing of AI itself assumes concentration in the model — a design flaw that should be remedied via superior infrastructure for federated oversight, federated training, and federated inference.
More from AGI Musings
- AI advances causing burnout: taking a break to cool down — DeryaTR_ · 2026-09-02
- Betting AI Favors Defense in All Threats Is Wishful Thinking — ronbodkin · 2026-09-02
- Using interpretability probes as privacy-preserving monitors to check models without seeing outputs — anpaure · 2026-09-02
- As models commoditize, operational context and evaluations become the real moat — bigdata · 2026-09-02
- Discussion on Dual-Use Risks of Interpretability Research — aryaman2020 · 2026-09-02
- Justin Johnson on World Models and the Future of Spatial AI — CSProfKGD · 2026-09-02