I'm so confused. Why do the lawyers not simply check to see if the references are real or not? How hard is it to look through an LLM's output and do a quick search to see if any of the laws or cases mentioned in it are in fact unsubstantiated?
Lawyers blame ChatGPT for tricking them into citing bogus case law
141–150 of 180 posts
Re: Lawyers blame ChatGPT for tricking them into citing bogus case law
#142While the lawyers blamed ChatGPT, the totality of the circumstances seem to indicate that they're less than honest in doing so. There is a live-tweet of the hearing here: https://twitter.com/innercitypress/status/166683852676213965... , and you can follow along with the lawyerly cringe there. Okay, lawyer #1 (LoDuca, the one on the case in the first place) appears to have played essentially no role in the entire case…
https://twitter.com/innercitypress/status/166683852676213965... is worth reading in full.
I am not a lawyer, but I was cringing throughout.
For non-lawyers (like me), I found it helpful to check out the quote tweets.
For example, to explain how much of a cringe was "F.3d" - https://twitter.com/ReichlinMelnick/status/16668536501095424...
> OMG. This is the Chat-GPT lawyer. You learn what F.3d means in first year legal writing. To non-lawyers, this is a tiny bit like someone being asked "You know what the DOJ is, right" and getting something wild like "Directorate of Judges" as a response.
The correct meaning of "F.3d" is "Federal Reporter, 3rd Series".
And it is like a lawyer's worst nightmare to have an exchange with a judge like this:
> Judge Castel: Have you heard of the Federal Reporter?
> Schwartz: Yes.
> Judge Castel: That's a book, right?
> Schwartz: Correct.
The implication being that at bare minimum, the lawyer could have looked it up in a book. Like a first year law student would.
Re: Lawyers blame ChatGPT for tricking them into citing bogus case law
#143Earlier quoted context omitted.
OpenAI is clear about the fact that ChatGPT can hallucinate, where have they ever said otherwise?
This is what I don't get. Not only have they not said otherwise, but they put it right up front in a pretty easy to understand brief message before you start using it. I guess lawyers just click agree without reading too.
Re: Lawyers blame ChatGPT for tricking them into citing bogus case law
#144The cynic in me wonders if this isn’t part of a plan to create a legal precedent banning AI from handling legal disputes. Think about it: the legal profession is possibly one of the most threatened by the development of AI models. What better way to secure the professional future of the long tail of lawyers and paralegals?
Re: Lawyers blame ChatGPT for tricking them into citing bogus case law
#145Disbarment is appropriate here. Why not blame your laptop manufacturer for creating the hardware you used to file your fraudulent court documents?
Something interesting about the legal profession is that it is self-regulating. The state bars are typically not government organizations. Attorneys know that confidence in their profession is extremely important and they strike the balance between preserving that confidence and, you know, destroying someone's livelihood because they don't understand how LLMs work.
Re: Lawyers blame ChatGPT for tricking them into citing bogus case law
#146comments here are like "what dumb lawyers". Sure OK. But what can we say here about "GPT-4 passed the bar exam!" and how useful is that data, given that this does not imply GPT-4 has the actual skills of a human lawyer.
The lawyer here was probably not smart enough to distinguish between 3.5 and 4. Haven't seen anything to indicate this was the output of GPT-4
The authors existed. They do research in the area, so the title was very plausible.
But the paper didn't exist.
Re: Lawyers blame ChatGPT for tricking them into citing bogus case law
#147Earlier quoted context omitted.
It is literally just a goddamn language model. it is very good at making plausibly human-like sentences. It is not a general intelligence, it is not your friend, it is not a research assistant. It is not designed to deliver content which is correct , it is designed to deliver content which is similar to human language . It might get things correct most of the time! But that is purely incidental.
It is literally just a goddamn electric motor. it is very good at converting chemical energy into mechanical energy. It is not a universal engine, it is not your horse, it is not your servant. It is not designed to create movement which is correct , it is designed to create movement which is similar to a piston engine . It might move you forward most of the time! But that is purely incidental.
Re: Lawyers blame ChatGPT for tricking them into citing bogus case law
#148Earlier quoted context omitted.
The lawyer here was probably not smart enough to distinguish between 3.5 and 4. Haven't seen anything to indicate this was the output of GPT-4
oh ok! so if he used gpt-4, still might have been illegal but the output would have been perfect. good to know
Re: Lawyers blame ChatGPT for tricking them into citing bogus case law
#149Earlier quoted context omitted.
I think the case is getting the attention because it's not just some one spouting off online, it's a lawyer bumping into the legal system with false information that would otherwise be a massive legal no-no and they are trying to scape-goat it onto the new shiny software.
But they might have seen the shiny new software touted by "those smart AI guys as being revolutionary and passing the Bar!" and they don't hang out on HN all day so to them that's like someone saying "this bridge is sturdy!" and they walk over it without realizing they should really go over the nuts and bolts of it like a civil engineer to really be sure
It's not that they need to inspect every nut and bolt of the bridge, they just need to not walk over the bridge - or at least, not immediately start driving unreasonably heavy loads across it.
Re: Lawyers blame ChatGPT for tricking them into citing bogus case law
#150Earlier quoted context omitted.
But they might have seen the shiny new software touted by "those smart AI guys as being revolutionary and passing the Bar!" and they don't hang out on HN all day so to them that's like someone saying "this bridge is sturdy!" and they walk over it without realizing they should really go over the nuts and bolts of it like a civil engineer to really be sure
I just think idiots who touch fire should be burned - particularly lazy idiots whose high paid job in fact requires them to be extremely careful and precise in their actions, and who refuse to take responsibility for their actions afterwards. It's not that they need to inspect every nut and bolt of the bridge, they just need to not walk over the bridge - or at least, not immediately start driving unreasonably heavy l…
Are we talking about lawyers or the AI researchers?, because they certainly want to portray themselves as a modern day Prometheus