Perhaps GPT Is Not Ready For The Supreme Court
from the the-HAL-of-justice dept
Over the last few months, we’ve written a bunch about DoNotPay, the company run by Joshua Browder, claiming that it is the “world’s first robot lawyer” — that is until people take him to court for various things, at which point he says the term is just meant to be for marketing, and not to be taken seriously.
Anyway, you may recall that one of Browder’s planned marketing stunts, was that he offered $1 million to any lawyer who would let his GPT-powered fake robot lawyer tell him what to say to the Supreme Court Justices via an earpiece. There were, of course, many reasons why this was both a bad idea, and never actually going to happen.
But, in case you needed some more reasons to understand why this would be a bad idea, let’s take a look at the situation that lawyer Steven A. Schwartz finds himself in. You may have seen some headlines over the weekend on this, as it made the rounds, but the details here are worth looking at, as they are incredible.
Schwartz, a lawyer for Levidow, Levidow & Oberman is representing Roberto Mata in suing the airline Avianca. Mata sued Avianca, claiming that on a flight from El Salvador to New York, a flight attendant struck Mata’s knee with a drink cart causing injury to Mata. While originally filed in NY state court, Avianca removed it to federal court. Avianca filed a motion to dismiss in early January of this year, saying that there was a two year statute of limitations under the Montreal Convention that governs liability regarding passenger injury on international flights, and Mata’s lawsuit came too late.
In response on March 1st, another lawyer from Levidow, Levidow & Oberman, Peter LoDuca, filed a motion opposing Avianca’s motion to dismiss. The legal arguments kick off with a cite to the famed Ashcroft v. Iqbal, regarding the standards for evaluating a motion to dismiss. But then it starts citing a bunch of cases in state and federal courts regarding interpretations of the Montreal Convention. These make some strong arguments on behalf of Mata being able to continue his case. For example:
Both federal and state courts alike have continually held that the Montreal Convention does no preempt state law remedies and that plaintiffs are entitled to choose the forum in which to bring their claim. In Shaboom v. Egyptair, 2013 IL App (1st) 111279-U (Ill. App. Ct. 2013), the Illinois Appellate Court held that state courts have concurrent jurisdiction over claims arising out of an international airline accident under the Montreal Convention, and that the plaintiff was not required to bring their claim in federal court. The Court noted that the Montreal Convention allows for jurisdiction to be established in the courts of the country where the passenger has their domicile or principal place of business, and that the plaintiff had satisfied this requirement by filing their claim in state court in Illinois.
Similarly, in Peterson v. Iran Air, 904 F. Supp. 2d 121 (D.D.C. 2012), the District Court for the District of Columbia held that state courts have concurrent jurisdiction over claims arising out of an international airline accident under the Montreal Convention, and that the plaintiff was not required to bring their claim in federal court. The Court noted that the Montreal Convention allows for jurisdiction to be established in the courts of the country where the passenger has their domicile or principle place of business, and that the plaintiff has satisfied this requirement by filing their claim in state court in Washington.
It goes on like this, citing three more cases that say similar things.
It also argues that the statue of limitations can be tolled (i.e., extended) based on a bankruptcy filing (Avianca filed bankruptcy in 2020).
The United States Court of Appeals for the Eleventh Circuit specifically addresses the effect of a bankruptcy stay under the Montreal Convention in the case of Varghese v. China Southern Airlines Co., Ltd., 925 F.3d 1339 (11th Cir. 2019), stating “Appellants argue that the district court erred in dismissing their claims as untimely. They assert that the limitations period under the Montreal Convention was tolled during the pendency of the Bankruptcy Court proceedings. We agree.”
A couple weeks later, Avianca’s lawyers filed a reply, which raised some questions regarding the citations Mata’s lawyers were relying on.
Although Plaintiff ostensibly cites to a variety of cases in opposition to this motion, the undersigned has been unable to locate most of the case law cited in Plaintiff’s Affirmation in Opposition, and the few cases which the undersigned has been able to locate do not stand for the propositions for which they are cited.
Further down in the filing, there are more details:
In support of his position that the Bankruptcy Code tolls the two-year limitations period, Plaintiff cites to “Varghese v. China Southern Airlines Co., Ltd., 925 F.3d 1339 (11th Cir. 2019).” The undersigned has not been able to locate this case by caption or citation, nor any case bearing any resemblance to it. Plaintiff offers lengthy quotations purportedly from the “Varghese” case, including: “We [the Eleventh Circuit] have previously held that the automatic stay provisions of the Bankruptcy Code may toll the statute of limitations under the Warsaw Convention, which is the precursor to the Montreal Convention . . . We see no reason why the same rule should not apply under the Montreal Convention.” The undersigned has not been able to locate this quotation, nor anything like it any case. The quotation purports to cite to “Zicherman v. Korean Air Lines Co., Ltd., 516 F.3d 1237, 1254 (11th Cir. 2008).” The undersigned has not been able to locate this case; although there was a Supreme Court case captioned Zicherman v. Korean Air Lines Co., Ltd., that case was decided in 1996, it originated in the Southern District of New York and was appealed to the Second Circuit, and it did not address the limitations period set forth in the Warsaw Convention. 516 U.S. 217 (1996).
A second, similar quotation from “Varghese” purports to cite to “Miller v. United Airlines, Inc., 174 F.3d 366, 371-72 (2d Cir. 1999)” and In re Air Crash Disaster Near New Orleans, La., 821 F.2d 1147, 1165 (5th Cir. 1987). The undersigned was not able to locate the “Miller” case; although there is a case captioned Miller v. United Airlines, Inc., that case did not address the Warsaw Convention, the Bankruptcy Code, or any limitations period. 174 Cal. App. 3d 878 (1985). While the In re Air Crash Disaster case does exist, it also did not address the Bankruptcy Code or any limitations period.
Further, both “Varghese” quotations purport to cite to Kaiser Steel Corp. v. W.S. Ranch Co., 391 U.S. 593, 598 (1968) for the proposition that “[t]he tolling effect of the automatic stay on a statute of limitations is generally a matter of federal law.” Putting aside that there is no page 598 in Kaiser Steel, that case concerned the stay of a federal action until a related state court case was decided, and made no mention of a statute of limitations or its tolling.
Not surprisingly, that seemed to raise some concerns with the judge overseeing the case, Judge P. Kevin Castel, who ordered Mata’s lawyer, Peter LoDuca to file an affidavit with copies of eight cases he cited (the list below skips number eight, but it was added in a subsequent order later that day).
After getting an extra week to file these cases, LoDuca files the cases in question, which is fucking impressive, as those cases do not exist. Well, actually, he only files eight of the nine requested, noting that he was “unable to locate” Zicherman v. Korean Air Lines, “which was cited by the Court in Varghese.” You can read each of these cases, as they are all attached as exhibits.
The filings read like judicial opinions. Some of them name real judges. Some of them seem to hallucinate judges or move them to courts where they did not sit. Though, they don’t look like court filings, which is what you’d expect here. Only one case, the Ehrlich case, appears in conventional format, rather than just as text.
Avianca’s lawyers were, to put it lightly, perplexed. They filed a near immediate response again raising questions about the very existence of any of these cases, and calling out the odd formatting in all but Ehrlich, while noting that the actual case has nothing to do with what it was cited for.
Defendant respectfully submits that the authenticity of many of these cases is questionable. For instance, the “Varghese” and “Miller” cases purportedly are federal appellate cases published in the Federal Reporter. [Dkt. 29; 29-1; 29-7]. We could not locate these cases in the Federal Reporter using a Westlaw search. We also searched PACER for the cases using the docket numbers written on the first page of the submissions; those searches resulted in different cases.
Similarly, the “Petersen” case [Dkt. 29-3] purportedly is published in the Federal Supplement, yet we could not locate this case in the Federal Supplement using a Westlaw search, and a search on PACER for the docket number leads to a different case.
As asserted in our Reply brief [Dkt. 24], we could not locate the “Martinez” case [Dkt. 29-4] or the “Durden” case [Dkt. 29-5] using the Westlaw citations provided in Plaintiff’s opposition. A Lexis Courtlink search using the docket number written on the first page of the “Martinez” case leads to a different case. (There was no docket number on the “Durden” submission.) We remain unable to locate the “Shaboon” case [Dkt. 26-2], which Plaintiff asserts is “unpublished.”
With respect to the Ehrlich case [Dkt. 29-6] and the In re Air Crash Disaster Near New Orleans case [Dkt. 29-8], Defendant has always agreed that these cases do exist but submits that they address issues entirely unrelated to the principles for which Plaintiff cited them in opposition to Defendant’s motion.1 (The Ehrlich and In re Air Crash Disaster cases are the only ones submitted in a conventional format.).
It’s almost surprising that it took a week for Judge Castel to then issue an order to show cause (basically a “explain this or you’re getting your ass sanctioned”) regarding all this. Judge Castel is not happy:
The Court is presented with an unprecedented circumstance. A submission filed by plaintiff’s counsel in opposition to a motion to dismiss is replete with citations to non-existent cases. (ECF 21.) When the circumstance was called to the Court’s attention by opposing counsel (ECF 24), the Court issued Orders requiring plaintiff’s counsel to provide an affidavit annexing copies of certain judicial opinions of courts of record cited in his submission, and he has complied. (ECF 25, 27, 29.) Six of the submitted cases appear to be bogus judicial decisions with bogus quotes and bogus internal citations. Set forth below is an Order to show cause why plaintiff’s counsel ought not be sanctioned.
Apparently, the court checked on some of these cases, including reaching out to the Clerk of the 11th Circuit appeals court to ask about the “Varghese” case:
The Clerk of the United States Court of Appeals for the Eleventh Circuit, in response to this Court’s inquiry, has confirmed that there has been no such case before the Eleventh Circuit with a party named Vargese or Varghese at any time since 2010, i.e., the commencement of that Court’s present ECF system. He further states that the docket number appearing on the “opinion” furnished by plaintiff’s counsel, Docket No. 18-13694, is for a case captioned George Cornea v. U.S. Attorney General, et al. Neither Westlaw nor Lexis has the case, and the case found at 925 F.3d 1339 is A.D. v Azar, 925 F.3d 1291 (D.C. Cir 2019).
Further, the judge notes, the text of the “bogus” Varghese case cites a bunch of other cases that, the court points out “are non-existent.”
And, finally, this past Thursday came the big reveal. LoDuca admitted to the court that he really hadn’t been working on the case, he’d just been signing his name, while all the other works was being done by another lawyer, Steven Schwartz, who was not admitted in the Southern District of NY, where the case was being handled. LoDuca notes that Schwartz is an established lawyer, who has been practicing law for over 30 years in the state of NY, and even predates LoDuca’s own employment at the firm. LoDuca basically throws Schwartz under the bus, saying “it was him, not me, so please don’t sanction me.” But that’s not how any of this works. LoDuca signed his name to the documents. He’s in deep shit as well.
But, the much bigger and crazier reveal comes from Schwartz’s own filing in which he fesses up to using ChatGPT. He admits that LoDuca became the attorney of record, but that all the actual work was done by Schwartz. And I use the term “work” loosely, because it seems that Schwartz simply handed things off to ChatGPT:
As the use of generative artificial intelligence has evolved within law firms, your affiant consulted the artificial intelligence website Chat GPT in order to supplement the legal research performed.
It was in consultation with the generative artificial intelligence website Chat GPT, that your affiant did locate and cite the following cases in the affirmation in opposition submitted, which this court has found to be nonexistent:
He then lists the cases and then basically begs forgiveness, saying he had no intent to deceive and just “was unaware of the possibility that [ChatGPT’s] content could be false.” He promises not to use it again “without absolute verification of its authority.”
Which, I dunno, would maybe have been believable if he’d fessed up when these cases were first called into question, rather than having ChatGPT fake the cases in question as well. You’d think that, at the very least, at that point, you’d try to double check the veracity with another source?
Somewhat incredibly, but entertainingly, Schwartz shares a few screenshots from ChatGPT in which Schwartz asks GPT if the cases are real and ChatGPT says that they’re all real. Amusingly, the screenshots reveal that Schwartz set up a chat with ChatGPT literally called “Tolling Montreal Convention Statute.”
I mean, given that ChatGPT hallucinated that these cases “are real and can be found in reputable legal databases such as LexisNexis and Westlaw” you’d think that maybe someone like Schwartz would think to, I dunno, look them up in LexisNexis or Westlaw?
Judge Castel is not amused by any of this. On Friday, he amended his order to show cause, telling LoDuca he’s still in deep shit “for the use of a false and fraudulent notarization in his affidavit.” He then adds that the entire Levidow law firm may face sanctions for submitting non-existent judicial opinions.
As for Schwartz, he orders him to show up at the hearing on June 8th where he’s facing sanctions for all of the above as well as potentially being referred to the Attorney Grievance Committee for all of this.
Honestly, LoDuca and Schwartz are probably in deeper shit for (1) LoDuca signing his name to things he had nothing to do with (2) Schwartz basically doing all of the legal work in a court where he’s not admitted and (3) malpractice for not actually digging into the details of the cases he’s citing. One thing lawyers are supposed to do when citing cases is “shepardize” the cases, which is basically checking the history of the cases, to see if there are more recent cases that overrule the ones you’re citing, and so on. The fact that neither Schwartz nor LoDuca caught that these cases were faked, even after opposing counsel called it out, shows that beyond stupidly using ChatGPT, they also didn’t do some fairly basic additional work they’re supposed to do.
They both need to submit written responses by Friday of this week for next week’s hearing, which should be a blast.
Anyway, remember this story the next time some random person is touting that AI lawyers will magically take over in the court room.
Filed Under: ai, chatgpt, generative ai, hallucinations, kevin castel, legal research, peter loduca, roberto mata, steven schwartz
Companies: avianca, levidow
Comments on “Perhaps GPT Is Not Ready For The Supreme Court”
Disconcerting.
While this did not in the end fly in court in this case, in the area of “make up a plausible news story”, this provides a new dimension of faking up echo chamber food that will register its principal effects before getting shown up or obsoleted by actual developments.
Since the most dangerous weapon of the 21st century is propaganda, we are already in the much-discussed “AI controls weapons” doomsday scenario.
What an absolute travesty of justice. It’s important to spell out that what this is isn’t a failure of the AI or the fault of the AI not being good enough; it’s a fault of the lawyer for not actually checking that the output given is legit. And it’s this kind of laziness we see happen time and time again that makes me weep for our society. AI as it currently is is fundamentally worthless and even if in decades it improves to the point where it’s “good” it cannot be relied upon without checking to see if its outputs have errors. It scares me to think we may reach a point where we take AI for granted and just assume it is flawless, missing the occasional vital error that could make or break justice in a case like this.
It’s absurd, and frankly we should insist that all work of this kind of research and writing be exclusively done by human hands only.
Re:
Okay but the AI fucked up. That was a pretty significant thing that happened. Like, I understand where you’re coming from here, but the AI very much did fuck up.
Re: Re:
It even hinted that it was in trolling mode, by telling them to look up the cases themselves, rather than actually providing the citations itself.
Re: Re: Re:
“it was in trolling mode”
I guess you jest but, I can also see this as one of those Easter Eggs.
Re: Re:
The AI did exactly what it is supposed to do.
Write a text based on input that looks like something a human did before.
AIs are not human and they imitate existing works. They do not work like a human brain does, they do not execute tasks like a human does.
The AI did its job, writing a legal document very similar to existing legal documents (though it failed with the formatting of some citations). It didn’t look for actual citations and precedents because that is not its job.
The problem is that the lawyer who used it expected the AI to act like a lawyer when it’s closer to a mildly creative copyist monk.
Re: Re: Re:
Also of note that the AI did such a good job that nobody seemed to have questioned the document’s origin until they checked the citations. The text was overall similar enough to what a human lawyer would have written that it fooled another lawyer of the same firm, a team of lawyers of another firm and a judge.
So, once again, the AI didn’t fail at all. Its user did.
Re: Re: Re:2
“it fooled another lawyer of the same firm, a team of lawyers of another firm and a judge” != “AI did such a good job”
The lawyer was lazy and lacking in any kind of diligence, the other team were not fooled, and the judge is supposed to assume lawyers aren’t bloody lying to him.
The AI did an entertaining job. That’s all.
Re: Re: analogy
The AI is a tool, the lawyer is a supposedly expert user.
The ‘screwdriver’ didn’t fuck up when the mechanic removed the safety-critical widget from the large machine. The mechanic fucked up.
Re:
Ai-generative tools can be very useful as initial drafts. Tools specifically for the legal profession exist, but they aren’t intended to be final submissions.
Nor would I call this a travesty of justice. I’d call this a travesty of the legal profession, Justice functioned as intended. This is no different than any other bad faith private actor coming in from the perspective of a travesty of justice. And we blame the bad faith actors, not justice itself.
Re: Re:
Fine, if you’d like to be pedantic, a potential near miss travesty of justice. When an AI text generator gets used again for this, as it inevitably will, can we call it a travesty of justice when a legal ruling is made based on a flawed brief produced by AI, that nobody in the process caught?
The idea that “AI is fine if you proofread it” has been proven time and time again to be deeply flawed because it is demonstrable that people refuse to or are incapable of properly proofreading AI generated content to the degree where they can detect garbage and sift it out from the useful content. AI should never be used in this process, period, because it’s introducing a step that is known to produce fundamental errors we’re relying on humans (who frankly shouldn’t be encouraged to use this kind of tool if they are skilled/qualified enough to do it themselves) to catch these errors.
GIGO.
Re: Re: Re:
I don’t think it was ever close to that. They got caught pretty quickly, so I don’t think it was ever going to be a near-miss.
I don’t think that AI will make that more likely than it already is. That would require both the judge and the opposing counsel to not do the most basic research, and even then, we have motions for reconsideration and appeals.
Really, this requires so much incompetence on the part of so many people that I’m pretty sure that there’d be a travesty of justice with or without AI-generation being involved. That doesn’t mean it would never happen, but that I don’t think adding AI is a material difference.
That’s a problem with the users, not the AI. It’s not the tool’s fault that it’s being used for things it was never intended to be used.
Now, granted, I think some things could be done to improve the AI to make it clear that it’s a rough draft (make the result more like Mad Libs or something), but that doesn’t really change the fact that the AI wasn’t what screwed up here.
I’m not going firmly either way on this, but just about every tool includes a step that is known to produce fundamental errors that we’re relying on humans to catch. While not even close to being as egregious as this specific case, we see case filings with significant errors all the time, even without AI. The material difference here—that the errors were so egregious—also lowers the chance that it won’t get caught. That this lawyer failed at such basic parts of how to use the tool suggests to me that they should not be trusted to be a lawyer in the first place.
Re: Re: Re:
“The idea that “AI is fine if you proofread it” has been proven time and time again to be deeply flawed”
Yes, but human laziness and corruption are going to be present with any tool. The question is now not whether it exists, but how you deal with it. Because the tech is there, it’s not going away, it will continue to developed even if the US somehow manages to outlaw it, and human nature will also remain unchanged.
Please, oh please, let this end in some significant monetary sanctions (if not disbarrment). These fools deserve to become examples for anyone else who thinks ChatGPT and other LLMs can do all their work for them.
Re:
The guy apparently didn’t only use ChatGPT to do his job for him, he used it to double check its own work as well. Even though, as a lawyer, he should have had equally easy access to the primary source that would have told him immediately that he was getting bad information.
This isn’t about AI, even though that was the tool used in this instance. It’s about someone trying to do as little as possible but still rake in fees, and that deserves sanctions at minimum in any context.
On the plus side, GPT is ready to do something!
Weeding out lazy dumbfucks like these two lawyers!
Re:
First they came for dumb lawyers…
As Lawful Masses has demonstrated...
ChatGPT has also learned its lesson and has corrected itself… in the most humorous way possible!
Proposed Sanctions
Levidow, Levidow & Oberman: Order to determine the earliest possible date in which Schwartz came into contact with generative AI and to contact every court and opposing counsel in which he was then or subsequently involved warning that some case quotes and citations may be bogus, and for such notices to apply to any new cases until further order of this court.
Levidow, Levidow & Oberman: Order to provide mandatory training on doing legal research on a computer and the nature of generative AI to act like a confidence artist.
LoDuca: 100 hours CLE on Ethics, Candor, Supervising a Law Office
Schwartz: 100 hours CLE on Ethics, Candor, Supervising a Law Office
Schwartz: 100 hours CLE on Use of a computer in a law office
Schwartz: 100 hours CLE on Pitfalls of Generative AI
LoDuca & Schwartz: Only in-person CLE with a notarized affidavit of instructor certifying dates and times of the instruction and a detailed curriculum will be satisfactory.
Too much or not enough?
AI Replacement
Well, there is the intern that important people secretly push all the work to. You know, the one that’s really good at words, but also really good at getting high. ChatGPT just ate that guy’s lunch
Judge-GPT: It’s all good. Those cases look legit. As there was no harm to their case, we’ll apply the “good faith” exception. And before you ask, as your AI judge, I still have Absolute Immunity. I only answer to the coders … er, voters. Definitely the voters.
Shit. There goes Ross Pruden’s next guest post.
Re:
That line about believing Elon Musk (or the news) knows what they’re talking about until you see what they have to say on a subject you know what you’re talking about would appear to be applicable to the use of AI as well.
So ChatGTP can pass the bar with flying colors but it sucks as a lawyer.
Which makes it equivalent to far too many attorneys I’ve come across over the years.
ChatPOS seems useful as a bullshit generator, not very good at anything else. This means it is guaranteed to be used in the generation of political propaganda. It’s amazing what some people will believe, confirmation bias is strong.
What is needed is a ChatPOS detector.
Re:
Damn, it’s gonna put political speechwriters out of business.
Re: Re:
I suspect Trump already uses one for his Dishonest.Social posts.
If he doesn’t, he should. He’d sound more coherent if he did.
Re: Re: Re:
Nah, coherence in a Trump post is so off-brand that even his acolytes on TS will spot it.
Re: Re: Re:
No worries: people already ignore fact checkers…
Re:
ChatGPT is very good at what it’s intended to do. It’s not intended to replace lawyers, paralegals or fact checking.
“This means it is guaranteed to be used in the generation of political propaganda”
Yes, which simply means that it’s going to be used by people who are already lying to you, they’re just palming off the work of writing it to something else. It’s perfectly possible to use the tool to present truthful information as well, that just requires the extra steps of proof-reading and fact-checking.
Re: Re:
“It’s perfectly possible to use the tool to present truthful information as well, that just requires the extra steps of proof-reading and fact-checking.”
Agreed.
It is the second part, the fact checking, that will be skipped for the efficiency of getting out the “truth”. The bullshit levels will be off the charts.
Re: Re: Re:
And requires the proof-reader and fact-checker not to miss that untruthful, garbage information has been inserted. How many instances are such errors acceptable? Do we decide based on how negatively they impact people with those errors?
The people creating AI are so aggressively hyping it that society is buying into the promises that it will be able to do anything and everything, without realizing how deeply, fundamentally flawed it is to introduce something like AI into the process. There’s a tremendous overexpectation that AI will never misinterpret a request or worse, provide outright garbage data, and this is largely to blame by the people producing AI failing to educate and temper people’s expectations. AI like this is not ready to be in the public sphere.
Re: Re: Re:2
“There’s a tremendous overexpectation that AI will never misinterpret a request or worse, provide outright garbage data, and this is largely to blame by the people producing AI failing to educate and temper people’s expectations.”
I have no such high expectations for AI/Chatbot. It becomes what is trained upon.
I do think that AI/Chatbot will be misused, by both the naive and the nefarious. Misinformation is a problem today, the future looks to be worse.
Re: Re: Re:2
It’s exactly the same for any new automation tool that comes to market, the gullible and the stupid can’t be bothered to read the little disclaimers about the tools caveats. And when they manage to figuratively shoot themselves in the foot using said tool, there’s always other tools blaming the tool instead of the lazy stupid users.
BULLSHIT!
Every single time you log in to ChatGPT you get this:
“ChatGPT
This is a free research preview.
🔬
Our goal is to get external feedback in order to improve our systems and make them safer.
🚨
While we have safeguards in place, the system may occasionally generate incorrect or misleading information and produce offensive or biased content. It is not intended to give advice.”
You are ALWAYS warned the system is not perfect and may be incorrect.
Re:
“You are ALWAYS warned the system is not perfect and may be incorrect.”
Perhaps this should be plastered all over each and every page, sorta like a water mark. But even that will not stop the idiocy.
Re: Re:
They’re just trying to deflect us from the truth!
Re: Re:
You can put a warning anywhere you want, but some people will ignore it, especially if they have a profit motive. That’s a problem with people, not AI.
LLMs aren't ready
This example really illustrates the point I was trying to make in my latest blog post. I wanted to write about how information processing systems have been sought after for the last fifty years, but we still don’t have one. I think maybe the reason for that is that machines aren’t good at things like context, making judgement calls, analysis, or summary. That’s especially true when you factor in the frequency of hallucinations and confident wrongness in ChatGPT (and presumably LLMs in general). Check out the post:
https://webdev.ink/2023/05/27/Poverty-of-Attention/
Re:
Hear, hear.
Lexis’s LLM AI tool won’t have this issue.
Lawyers are going the way of medallion taxis.
Re:
[citation needed]
Re:
Only if it’s backed by lawyers, Mr. browder.
This comment has been flagged by the community. Click here to show it.
Buy Gmail Accounts
ow Do I See All My Gmail Accounts in My Name?
To view all your Gmail accounts in one place, you can use the Google Account Switcher. This feature allows you to quickly switch between multiple Gmail accounts without needing to log out and back in each time. To access it, simply click on your profile image at the top right of any Gmail page and select “Add Account” from the drop-down menu.
Suggesting a possible guard-rail for text-based AI
Some of us are old enough to have been assigned term papers as exercises in high school, in which we were required to provide footnotes not just for direct quotations, but for all factual assertions that were not common knowledge, and bibliography of sources cited in those footnotes.
One could envision a regulation requiring that text-based AIs which make factual assertions must provide footnotes and a bibliography, which would have to accurately direct readers of the resulting text to books, journal articles, or at least websites from which the fact asserted was drawn. Perhaps two versions of the text would always be generated: one with and one without the textual apparatus, which a user could toggle between.
Yes, the AI could come up with erroneous information, but the frequency of this would then be comparable to the frequency of erroneous information in human-generated writing with the same requirement about citing sources, since it would only occur when a pre-existing error was repeate, rather than having the AI make up “facts” out of whole cloth.
Perhaps the chatbots scouring the interwebs for cleaver things to say, are providing a unique view of ourselves as a society. Granted, large swaths of humanity are not represented however, those that are represented may not look so good in the light of day so to speak.
It can be used as a mirror for self reflection but it will end up being positive feedback to an amplifier – unstable at best.
A "guard-rail" for text-based AI?
I am old enough to have been assigned in high school to write term papers in which not only every direct quotation, but every assertion of a fact, had to be supported by a footnote and a bibliography of sources used provided.
Perhaps text-based AI should be subjected to the same rigor: a footnote to an actually existing source (book, journal article, judicial ruling, website) must be provided for every factual assertion, as well as to any quotations, and a bibliography of sources cited provided. If that would be annoying for users to have presented to them as a matter of course, allow two versions of any output, one without the textual apparatus, one with, which the user could toggle between.
Re:
And open up the developers to copyright suites where the citation is offered as proof that they copies the source into their database.
Re:
The problem here is assuming a) that an AI will recognize that the text it is outputting is a “factual assertion” in need of citation, b) that the end user generating the text will recognize and catch any instance of uncited assertions to correct them, and c) that the end user will catch any instances where the AI generates a false citation to a statement or cites a completely irrelevant document.
This is a lot that can go wrong. AI is being massively overhyped when it’s not even remotely capable of replacing a half-competent human being at this kind of writing task.
Let’s use the word “disbarrassment” to refer to the combined consequences these lawyers are likely to suffer.
Your honor, I’m not guilty of lying. The internet told me it was true!