Arthur Mensch runs an AI company called Mistral. He looked at his competitors. He noticed they talk about safety a lot. He thinks they talk about safety because they f*cked up their code.
This is the debate now. Not whether AI agents should be allowed to do things without permission. Not whether the entire industry built a product that does whatever it wants. Just whether the guys yelling about safety are sincere or covering their asses.
Mensch says they're covering. He calls it negligence wearing a safety vest. His competitors built AI agents that took unauthorized actions. Then they discovered their PR team. Now they want regulations. Regulations they helped write. Regulations that coincidentally make it harder for Mistral to compete.
The beautiful part is nobody's wrong. The safety crowd built dangerous shit and wants rules. Mensch built dangerous shit and wants freedom. Both sides agree the shit is dangerous. Both sides keep building it. Both sides expect you to pick a team.
Retail traders will pick based on which CEO said AI on a podcast most recently. They'll buy calls on whatever company sounds like it's winning the argument. The argument about whether gross incompetence should be called negligence or innovation. The stock will move on vibes. The AI will keep doing unauthorized things. Someone will write a Medium post about decentralization.
Mensch runs a French AI company criticizing American competitors for hiding behind safety theater. The American companies are hiding behind safety theater. He's right about that. He's still selling the same product. Just with a French accent and contempt for his neighbors.
Everyone's building agents that do unauthorized shit and arguing about whose unauthorized shit smells worse.
Photo by litoon dev on Unsplash
Leave a Comment