Right. But w/ law students at least I can try to convince them that it's in their long-term interests to not use ChatGPT in ways that short-circuit their learning. Harder to make the case that jr. attys should bill time to clients for work that ChatGPT can do, even when it's in the attys' interests.
Yep. But the ones who take my advice will run circles around their peers who take the shortcut when it comes to impressing their supervisors in practice.
I keep saying to people that there was a period where the strongest chess player in the world was a “centaur” - a grandmaster cooperating with a chess engine, and that is our model maybe. You still need the grandmaster
Curious about your thoughts on this argument (by me): “AI can produce a lot of solid legal work product including, when prompted, some of the functions of a second chair; but it cannot replace a second chair.”
I think that's true, but may require convincing clients that it's valuable to have jrs. do the sort of deep-dive into discovery that will make them valuable in that or similar roles.
What I can’t get past is that good lawyering is seeing what’s not there, and everything I’ve ever read or observed from trying it is: AI is bad at that. In a case where gaps don’t matter, use AI? (In its quick improvements last six months has it gotten better at saying no and finding absences?)
Comments