Feels like overlooking the same issue as with every other AI use

When a human makes a mistake and is called out, they can usually fix the mistake. When genAI outputs nonsense, it’s fucking nonsense, you can’t fix something that’s fundamentally made up, and if you try to “ask it” to fix it it’ll just respond with more nonsense. I hallucinated this case? Certainly! Here’s 3 other cases you could cite instead: 3 new made up cases

So if you knew the average lawyer made 3.6 mistakes per case and the AI only made 1.2, it’s still a net gain.

thats-not-how-any-of-this-works.webm

It’s hard to get into the article’s mood when I know that Lexis not only still exists but is now part of the Elsevier family; this is far from the worst thing that attorneys choose to do to themselves and others. Lawyers have been caught using OpenAI products in court filings and court appearances, and they have been punished accordingly; the legal profession does not seem prepared to let a “few hallucinated citations go overlooked,” to quote the article’s talking head.

They will ignore your errors and respond with errors of their own. AI will decide you’re guilty and deny your appeal.

A later case exactly like yours will result in an innocent verdict because the case used one different word and the butterfly effect will cause the AI to add the word “not” to the verdict.

AI will conclude that your case was ruled in error, but there’s nothing it can do because the appeal was already denied

Yeah, I don’t care about the raw amount of mistakes, I care whether the mistakes are severe enough to throw the case. Stuff like missing filing deadlines.

Except nobody wants to tali about that 1.2 being it thinks green is an object or something completely fucked.