It’s kind of insane that we expect AI entering a human world to meet a 100% safety standard. While system is already significantly better than humans, but it’s not a God humans crash every single day and make a lot of mistakes Humans drive drunk, tired, distracted, emotional and looking at their phones Most of those crashes barely make the news One minor Tesla crash and it’s everywhere. The coverage is totally different Tesla FSD is already significantly safer than humans Tesla’s latest safety data shows FSD is roughly 7x fewer collisions than the U.S. average That’s an 86% lower collision rate. That’s not a small improvement. That’s an enormous safety difference and real lives saved The question was never: “Can AI ever crash?” It should be: “Can AI crash dramatically less than humans?” Tesla already showed the answer: 7x better Autonomy doesn't need to be perfect to save lives... (while in fact it will keep getting better) It just needs to be dramatically better than us