He may want to read up on how many lawyers have already lost their license due to relying on AI, because you know, AI just makes shit up from time to time. Most competent lawyers are avoiding it like the plague now. We tried implementing it in a handful of ways to cut down on doc prep and you end up spending more time than before running back through everything and correcting all of the errors. The things that are simply missed, rather than the glaring mistakes, are the biggest concern as they're harder to catch. We scrapped all of it, and rightfully so. AI is fun toy, and has some applications where it can be beneficial. Anything that involves precision shouldn't be anywhere near it. Mostly it's just a vehicle for violating IP rights, as well as privacy and publicity rights. So the Internet and children love it, naturally, but I'm still waiting for real world applications that actually amount to anything of substance. The one area you can maybe point to, the applications in the discovery of new antibiotics and other drugs leave me skeptical, frankly. Given how shitty it is at everything else, should we really be trusting what it outputs for pharmaceuticals?