Debate on AI Safety: Concentration of Power Not Inherently Bad
sethlazar · x · 2026-09-01
Amidst the AI safety community's focus on "concentration of power," the author offers a critical perspective.
Core Arguments:
- Concentration of power is not inherently bad. Liberal democratic states, parents, and CEOs all hold power without it being automatically negative.
- Criticism should move beyond merely naming power to examining its justified use and procedural legitimacy.
- Debates must clearly define what constitutes good power, just exercises of it, and where to draw the line between acceptable and unacceptable concentrations.
The author emphasizes the need to make a clear case for individual liberty, distinguishing it from other configurations of society and power.
Related event: AI Safety Community Debats Power Concentration(4 posts)→
More from AGI Musings
- LeCun, Hassabis, and Li Converge on World Models as AI's Future — TheTuringPost · 2026-09-02
- Steve Yegge: All models will eventually build systems they can't maintain — Steve_Yegge · 2026-09-02
- Gary Marcus on reality vs satire: GoPro pivot to AI data centers sends stock up 38% — GaryMarcus · 2026-09-02
- Thom Wolf: Future Interfaces Will Use Live Diffusion, Software Will Use LLMs — c_valenzuelab · 2026-09-02
- Ken Liu, author behind Pantheon, publishes essay "The Art of Copying" on AI — avilacjf · 2026-09-02
- Sam Altman: Superintelligence May Happen Sooner Than Thought; OpenAI Will Build Humanoid Robots — ChrisGPT · 2026-09-02