Guillaume Verdon

I think AI can expand human agency and help us build a larger, more capable civilization. The best future is not one centralized intelligence deciding for everyone, but personal superintelligence augmenting each person, alongside competition among models, companies, ideas and technological approaches. That diversity matters because adaptation requires variation. Concentrated control creates brittle systems and suppresses experimentation. AI also shifts the frontier of intelligence. Virtual reasoning and mathematical discovery can advance much faster than intelligence embodied in the physical world, where robotics still confronts messy reality. If recursive self-improvement genuinely works, then access to compute could become increasingly decisive—but that is a conditional argument, not a law of nature. My broader thermodynamic framing is an interpretive lens: intelligence helps civilization identify possibilities, marshal energy and resources, and create new forms of order. It does not prove that every AI deployment is good. There will be harms and failures. I favor holding actors liable for concrete damages, which creates adaptive incentives to improve safety, rather than constructing a centralized bureaucracy around speculative catastrophe. The goal should be to accelerate capability while preserving open competition, distributed power and room for civilization to evolve.