Should AI companies be allowed to write the safety rules that will govern them?
When the lab that insists AI developers can’t grade their own homework drafts the federal brake meant to stop it, with the trigger set above a billion dollars in damages or fifty deaths and the inspectors paid by the inspected, is industry-written regulation a pragmatic head start or a conflict of interest that a democracy shouldn’t accept?
Commentaires (1)
In today’s episode of Minds, Bodies, and Terawatts (October 19, 2026), the hosts dig into POLITICO’s report that Anthropic quietly pitched Congress a confidential draft AI safety bill of its own making. The show frames it as a circuit breaker wired by the people it is meant to trip, and asks whether a threshold rated above every incident we’ve actually seen is a brake at all. The aviation comparison does a lot of work: Boeing doesn’t pay the people who investigate its crashes, and a century ago America learned why crash investigation couldn’t live inside the agency that certified the planes. There is a real argument on the other side too, since a bill written by people who understand the technology might beat one written by people who don’t. Listen to the full episode and tell us where you’d set the brake, and who should hold the pen.
Related reading on unscarcity.ai:
Envie d'aller plus loin ?
Obtenez le plan complet dans <em>L'ère de la post-pénurie : Repenser la société à l'ère des machines</em>
