Hacker News Reader

Comment by sgbeal | Hacker News Reader

That should be a career-ending move, not a fine.

I disagree with the notion that it should be "career ending", but I do think they got off pretty light.

They doubled down on their original mistake rather than admit what they did, even when it was quite clear the judge and opposing council was on to them. They should be disbarred.

Where is this doubling down? It certainly isn't in this article.

See https://storage.courtlistener.com/recap/gov.uscourts.nysd.57... - Page 2, the judge uses the very words "doubled down": "But if the matter had ended with Respondents coming clean about their actions shortly after they received the defendant’s March 15 brief questioning the existence of the cases, or after they reviewed the Court’s Orders of April 11 and 12 requiring production of the cases, the record now would look quite different. Instead, the individual Respondents doubled down and did not begin to dribble out the truth until May 25, after the Court issued an Order to Show Cause why one of the individual Respondents ought not be sanctioned."

I guess that really depends on what you consider doubling down.

-On March 1, They filed a submission with fake citations from chatGPT.

-On April 11th court told them the cases couldnt be found and to produce full transcripts of the cases by April 25.

-On April 25, They came back and said they don't have them and couldnt find the cases, only the limited excerpts provided in "online databases".

-On May 4, the court asked them why they ought not be sanctioned for citing non-existent cases to the Court.

-On May 25, They said we found these cases on ChatGPT and thought they were real.

I guess the more transparent response on April 25 would have been to say "We weren't able to find the full cases and we now believe they probably dont exist ". Im not sure if the latter part is material if they already admitted they cant find the cases. If you cant produce the cases, is the reason relevant at that point?

> They said we found these cases on ChatGPT and thought they were real

Just like Wikipedia, ChatGPT is great for generating leads, but you need to check the actual primary sources! How do people not get this?

>Where is this doubling down? It certainly isn't in this article

It is mentioned also in the article:

>then continued to stand by the fake opinions after judicial orders called their existence into question.”

It is:

> then continued to stand by the fake opinions after judicial orders called their existence into question.

Did they?

Here’s a longer video with more explanation: https://youtube.com/watch?v=oqSYljRYDEM&pp=ygUOQ2hhdGdwdCBsY...

At first it was just generated citations. The opposing lawyers and judges could not find the cases, and so they then asked chatGPT for them, and it generated them, and they were submitted to the court. (One of the lawyers notarized the other’s filings.)

There are even transcripts of chatGPT telling the accused that it is not an authoritative or up-to-date legal database.

—-

The punishment could arguably be more severe, since it demonstrated that these lawyers do not read their cases, particularly the one that notarized everything.

They should be disbarred.

I just can't agree. To my way of thinking there are very few (if any) things that should be "career ending" the first time they happen. And the main guy involved here, as I recall, has a long career with what has been a spotless record until now. I just can't see the justification for ending that career over one fuck-up.

But, again... the punishment does seem lighter than I expected. I was thinking that any fine associated with this should be enough to hurt a bit. Something like $250,000 or something. This amounts to a slap on the wrist.

Just to clarify my position here: I do think this is a major fuck-up. Like, really major. To the extent that I think the punishment should not be "career ending" it's just a reflection of my seeing "career ending" punishments as something of a "nuclear option" that should be avoided in all but the absolute most extreme cases. I guess I would say that if punishments spanned a scale from 1-10 with "10" being "disbarment" then this should probably be about an 8 or 9 in my mind.

An officer of the court notarized that they had reviewed the imaginary citations and that those imaginary citations were material to the case. This is a gross breach of ethics and, given how easy it is to look up cases and verify them, grossly incompetent and lazy.

These people should not be practicing law.

The judge cannot disbar them, but the opinion provides all of the evidence necessary.

The punishment isn’t just the $5,000, it’s the public shaming that came with it. You’re not going to be able anyone’s name here for many years without this case coming up. The world at large has been watching this unfold gleefully.

That, combined with the lawyers involved having decades-long clean slates, as well as this case being outside the normal realm this team would operate in, as well as ChatGPT being relatively new at the time this happened makes it all seem pretty proportional.

Maybe you could say they deserve worse for how they handled things after they got caught, but I struggle to really have a problem with this punishment considering the records of those involved.

> The punishment isn’t just the $5,000, it’s the public shaming that came with it.

Both the fine and the public shaming are very light "punishments". Everyone's pointing and laughing now, but in a year nobody will even remember their name and their career will be fine.

local legal communities have long memories. they’ll be griefed about it at their retirement parties.

> local legal communities have long memories.

So does Google, and anyone researching these lawyers or this firm in the future...

Sure, but when you Google a lawyer's name and this comes up you may choose not to use him.

On the other hand, there are certain kinds of lawyer with low-research clients who won't suffer much from this, and these lawyers are likely that kind (representing a client suing Avianca for a drinks cart complaint).

Why so eager to end someone's career? He did a lazy/stupid thing. People do far worse things.

The client may want to sue them and seek damages or something and that'd be fair but I don't think their career should end. That's harsh.

Things like this is what causes lawyers to lose cases for their clients. Imagine being a client of a lawyer who uses ChatGPT. You lose the case and then your livelihood or worse.

These twits should be disbarred, and without hesitation.

> Imagine being a client of a lawyer who uses ChatGPT. You lose the case and then your livelihood or worse.

This presupposes that the use of ChatGPT is the reason the case was lost. In this case, the judge found the mistake, called it out, and fined the lawyers for their carelessness. It doesn't necessarily follow that p1) ChatGPT was used, p2) Case was lost, c) Case would have been won if ChatGPT was not used.

No, the use of ChatGPT is the only reason that they thought they had a colorable argument that the case could be won.

> Things like this is what causes lawyers to lose cases for their clients.

Or to waste a lot of their clients money on cases that they should have known could not be won from the outset, as is the case here.

What they did was the professional equivalent of a surgeon thinking "I wonder what'd happen if I chopped out this thingie?", then acting on that impulse, then lying about his actions.

With that logic all us programmer's careers should be ended due to the bugs we introduce because "they are equivalent of a surgeon thinking...".

It's not equivalent to that. Nobody lost their leg over this.

Most programmers don't have to pass through rigorous education to be professionally certified on par with lawyers or surgeons.

Nobody cares about lazy/stupid. The problem is that he lied about it again and again.

Stupid mistakes are usually not career ending, fraud and cover up usually is.

Because to do otherwise is to normalize the use of false or non-existent information, whether due to bad faith, laziness, or gross ignorance, in processes where the outcomes may affect millions of dollars, thousands of people, or individuals' lives and freedoms.

Because if someone gets a minor fine, and doesn't lose their career, it's less clear why this would be fun to talk about. This is the most important dynamic in message board threads about current events.

> Why so eager to end someone's career?

Because certain careers, e.g. lawyers, surgeons, and law enforcement give their practitioners the ability to make or break their clients, potentially killing them or ruining them financially and/or socially. Such professions must, for reasons of social sanity, be held to higher standards than someone who's career is, e.g., writing web apps or serving coffee.

Gross negligence. Checking sources is a cornerstone of professions that are highly reliant on citing sources.

Or if not end career some real punishment like enough time (by USA standards couple decades) in prison.

Why do you think that would be better than a warning or a fine?

Because it wasn't just a simple mistake. It was such egregious professional negligence that it calls into question not only the basic competence of the attorney, but also their basic sense of ethics.

To add onto this - the lawyers failed to verify their resources! This is something that every lawyer carefully pays attention to before submitting anything.

The fact that they did no research on ChatGPT, and they didn't verify the output from ChatGPT, is pretty bad.

They're being punished more so on how they used their resources, not for using ChatGPT.

> They're being punished more so on how they used their resources, not for using ChatGPT

Yes, exactly. If they'd used a human intern instead of ChatGPT, it wouldn't change the situation at all.

Huh, I think it would? It's still really bad, the quality of the service the client got is still abysmal, and it's still ultimately the lawyer's responsibility. That much is certainly unchanged, it would also be considered a much more understandable mistake.

If they followed exactly the same processes (showing a screenshot of a text conversation saying "Is X a real case?" when challenged about the nonexistent cases) then I imagine they would have have been treated the same way. However they probably wouldn't have gone down that route, because they know "uhhhh I included those cases because my intern insisted they exist?" absolutely wouldn't fly. The difference is that they thought they could chance it because ChatGPT is a shiny, new thing that some people have a pretty mixed understanding of.

It needs to be a warning to the industry, which requires real consequences for those involved. A warning could encourage abuse by others - there’s little to lose except having to stop after getting caught.

Reply on news.ycombinator.com