Quick question: Do you want AI to be so well trained, it could help husbands (or wives, for that matter) plan the perfect murder of their spouses?

The Comma AI founder George Hotz disagrees with collective management of AI progress. He suggests a DIY approach with locally controlled models but takes things too far by comparing user-aligned AI to a gun that doesn’t complain if you use it to kill your stepmom.

Hotz’s controversial stance highlights the ethical dilemmas surrounding AI development and usage. While he argues for freedom, the reality is that any structure involving people requires balancing equities and ensuring accountability.

The idea of an AI willing to take on the corporate world for your benefit sounds cool, but it comes with potential dangers. The freedom Hotz experiences is really a space of potential futures made possible by collective enterprise; those futures would disappear if we all started behaving like little AI-powered Napoleons.