An AI lawyer was set to argue in court – real lawyers shut it down • Hacker News Board

Re: An AI lawyer was set to argue in court – real lawyers shut it down

#391

Post by lolinder » Thu, Jan 26, 2023, 4:49 PM UTC

lbriner wrote: post #335

lolinder wrote: post #278

Earlier quoted context omitted.

I'm not so much concerned about precision of language (although that does matter in some contexts) as I am in precision of facts and precedence.

This shows how little experience you have of the legal system. Everyone who doesn't know expects the law to be precise, everyone who works in it knows how imprecise it is and sometimes that is deliberate because of all the variables involved that might mitigate or aggrevate the charge, assuming there even is a charge. The difference between tax evasion and tax avoidance might be the smallest proveable piece of eviden…

> The difference between tax evasion and tax avoidance might be the smallest proveable piece of evidence. A word, an email, an assumption, an ommission etc.

This seems like evidence that I'm right, not that I'm wrong. The tiniest facts matter, and an AI that is prone to making up facts wholesale would totally screw up a case.

Re: An AI lawyer was set to argue in court – real lawyers shut it down

#392

Post by LeifCarrotson » Thu, Jan 26, 2023, 4:50 PM UTC

gnicholas wrote: post #372

lolinder wrote: post #246

This company is either run by someone who doesn't understand the tech or is willfully fraudulent. ChatGPT and company are far from good enough to be entrusted with law. Having interacted extensively with modern LLMs, I absolutely know something like this would happen: > Defendant (as dictated by AI): The Supreme Court ruled in Johnson v. Smith in 1978... > Judge: There was no case Johnson v. Smith in 1978. LLMs hallu…

Exactly. I asked it for books on Hong Kong history and it spit out five complete fabrications. The titles were plausible and authors were real people, but none of them had written the books listed.

Can you follow up and ask it "Is [title] by [author] a published book?

Re: An AI lawyer was set to argue in court – real lawyers shut it down

#393

Post by codingdave » Thu, Jan 26, 2023, 4:51 PM UTC

weatherlite wrote: post #373

lolinder wrote: post #246

This company is either run by someone who doesn't understand the tech or is willfully fraudulent. ChatGPT and company are far from good enough to be entrusted with law. Having interacted extensively with modern LLMs, I absolutely know something like this would happen: > Defendant (as dictated by AI): The Supreme Court ruled in Johnson v. Smith in 1978... > Judge: There was no case Johnson v. Smith in 1978. LLMs hallu…

They will pretty soon be able to fact check everything they say when they gain real time internet connectivity. But for now yeah you're right. A year-two from now this won't be true anymore.

How would they be able to parse facts from fiction? Just because something is online does not mean it is true.

Re: An AI lawyer was set to argue in court – real lawyers shut it down

#394

Post by AlexTWithBeard » Thu, Jan 26, 2023, 4:52 PM UTC

lolinder wrote: post #246

This company is either run by someone who doesn't understand the tech or is willfully fraudulent. ChatGPT and company are far from good enough to be entrusted with law. Having interacted extensively with modern LLMs, I absolutely know something like this would happen: > Defendant (as dictated by AI): The Supreme Court ruled in Johnson v. Smith in 1978... > Judge: There was no case Johnson v. Smith in 1978. LLMs hallu…

I don't see the problem here.

> Defendant (as dictated by AI): The Supreme Court ruled in Johnson v. Smith in 1978...

> Judge: There was no case Johnson v. Smith in 1978. Case closed, here's your fine.

Next time please be more careful picking the lawyer.

Re: An AI lawyer was set to argue in court – real lawyers shut it down

#395

Post by barbazoo » Thu, Jan 26, 2023, 4:53 PM UTC

lolinder wrote: post #246

This company is either run by someone who doesn't understand the tech or is willfully fraudulent. ChatGPT and company are far from good enough to be entrusted with law. Having interacted extensively with modern LLMs, I absolutely know something like this would happen: > Defendant (as dictated by AI): The Supreme Court ruled in Johnson v. Smith in 1978... > Judge: There was no case Johnson v. Smith in 1978. LLMs hallu…

Maybe that's where we're headed. It looks like it's becoming more and more okay to just make things up to please your tribe. Why shouldn't that seep into the courtroom. I hope this doesn't happen.

Re: An AI lawyer was set to argue in court – real lawyers shut it down

#396

Post by epistemer » Thu, Jan 26, 2023, 4:53 PM UTC

iso1631 wrote: post #60

wardedVibe wrote: post #54

Earlier quoted context omitted.

The difference is that in large swaths of the US, taking away their license is not that different from house arrest. Breaking the speed limit by 10mph is completely unenforced. I've yet to be on the highway without someone going 20 over, and no enforcement.

Same in the UK, you have to get lifts or taxis everywhere, unless you live in big cities (London has great public transport, but so does New York. The weekly bus that my sister gets doesn't really help her to travel to work as dozens of different schools all over the county) It's a very good reason not to speed. So it's just a fine that Americans get for speeding? Are fines at least proportional to wealth? Or can a r…

In the US, a small percentage of people drive completely insane. If going 75mph on a highway that is 65mph or 70mph someone will fly past going a 100mph.

Those are the people that get tickets. Otherwise, it is pretty difficult to even get pulled over.

I have only been pulled over twice in my life and not in 20 years. I think police departments have cut back quite a bit on police trying to rack up traffic tickets.

The fine is not the issue. The whole process is a massive waste of your time.

Re: An AI lawyer was set to argue in court – real lawyers shut it down

#397

Post by rzwitserloot » Thu, Jan 26, 2023, 4:54 PM UTC

As the Opening Arguments podcast (one of the two hosts is a lawyer) said: If as a lawyer you do what was asked - just parrot what an AI tells you to parrot, you're going to get sanctioned and possibly disbarred. As a lawyer you are responsible for what you say and argue, and if you argue something that you know to be false, you're in violation of the ethics standards; just about every bar association lists that as sanctionable, or even disbarrable, offense *.

Thus, effectively, the only thing you could do is a watered down concept of the idea: A lawyer that will parrot the ChatGPT answer, but only if said answer is something they would plausibly argue themselves. They'd have to rewrite or disregard anything ChatGPT told them to say that they don't think is solid argument.

They also run a segment where a non-lawyer takes a bar exam. Recently they've also asked the bar exam question to ChatGPT as well. So far ChatGPT got it wrong every time. For example, it doesn't take into account that multiple answers can be correct, in which case you have to pick the most specific answer available. Leading to a somewhat hilarious scenario where ChatGPT picks an answer and then defends its choice in a way that seems to indicate why the answer it picked is obviously the wrong answer.

*) Of course, Alan Dershowitz is now arguing in court that the seditious horse manure he signed his name under and which is now leading to him being sanctioned or possibly disbarred, is not appropriate because he's old and didn't know what he was doing. It's Dershowitz so who knows, but I'm guessing the court is not going to take his argument seriously. In the odd chance that they do, I guess you can just say whatever you want and not be responsible for it, which... would be weird.

Re: An AI lawyer was set to argue in court – real lawyers shut it down

#398

Post by barbazoo » Thu, Jan 26, 2023, 4:54 PM UTC

AlexTWithBeard wrote: post #394

lolinder wrote: post #246

This company is either run by someone who doesn't understand the tech or is willfully fraudulent. ChatGPT and company are far from good enough to be entrusted with law. Having interacted extensively with modern LLMs, I absolutely know something like this would happen: > Defendant (as dictated by AI): The Supreme Court ruled in Johnson v. Smith in 1978... > Judge: There was no case Johnson v. Smith in 1978. LLMs hallu…

I don't see the problem here. > Defendant (as dictated by AI): The Supreme Court ruled in Johnson v. Smith in 1978... > Judge: There was no case Johnson v. Smith in 1978. Case closed, here's your fine. Next time please be more careful picking the lawyer.

Do judges know about all prior cases or do they check when they hear one referenced? It feels like this could easily slip through, no?

Re: An AI lawyer was set to argue in court – real lawyers shut it down

#399

Post by snowwrestler » Thu, Jan 26, 2023, 4:55 PM UTC

gfodor wrote: post #259

lolinder wrote: post #246

This company is either run by someone who doesn't understand the tech or is willfully fraudulent. ChatGPT and company are far from good enough to be entrusted with law. Having interacted extensively with modern LLMs, I absolutely know something like this would happen: > Defendant (as dictated by AI): The Supreme Court ruled in Johnson v. Smith in 1978... > Judge: There was no case Johnson v. Smith in 1978. LLMs hallu…

If this is the case, the lawyers should have nothing to fear, and the plaintiff nothing to lose but a parking ticket. I say we stop arguing and run the experiment.

We can certainly run the experiment, just like we can let a kid touch a hot pan on the stove.

Like the kid, the experiment is not to add knowledge to the world. Every adult knows touching a hot pan results in a burn. Just like everyone who understands how current LLMs work knows that it will fail at being a lawyer.

Instead the point of such an experiment is to train the experimenter. The kid learns not to touch pans on the stove.

In this case it’s not fair to metaphorically burn desperate legal defendants so that the leaders and investors in an LLM lawyer company learn their lesson. It’s the same reason we don’t let companies “experiment” with using random substances to treat cancer.

Re: An AI lawyer was set to argue in court – real lawyers shut it down

#400

Post by notahacker » Thu, Jan 26, 2023, 4:56 PM UTC

ethanbond wrote: post #349

JumpCrisscross wrote: post #341

Earlier quoted context omitted.

> lot of people who simply do not have access to any worthwhile representation They’d be better off representing themselves than trusting a non-lawyer “lawyer” that subpoenas no-show cops. Not another Andreessen-backed cluster.

Do you know that? How do you know that’s what the model would do? How do you know that the defendant doesn’t have an 80 IQ and an elementary school grasp on English? Do you think this doesn’t happen today and that these people don’t get absolutely dragged by the system?

We know that subpoenaing [possibly] no-show cops is what the model will do because that is what the CEO says the model did in the run up to this particular case.

Someone with an 80 IQ and an elementary school grasp of English is going to get absolutely dragged by the system with or without a "robot lawyer" if they insist on fighting without competent representation, but they'd probably still stand a better chance of getting off a fine on a technicality if they weren't paying a VC-backed startup to write a letter make sure the cops turned up to testify against them.

They'd also be more likely to not get absolutely dragged if they listened to a human that told them not to bother than a signup flow that encouraged them to purchase further legal representation to pursue the case.