I’m on vacation this week and next, so this will be mercifully brief, I promise.
Search Engines and AI
First, there’s this great article on search engines and AI by Ryan Broderick. If you don’t subscribe to Garbage Day, you should.
Another Fake Cases, um, Case
Second, you’ve probably seen this on the LinkedIn hot take factory already, but there’s yet another “guy gets sanctioned for citing fake case law from ChatGPT” case. But this one’s a bit different in my opinion:
Karlen, in his response, confessed that he had hired a legal consultant in California who promised to help for a fee that was less than 1% of what it would have cost him to hire a Missouri attorney.
“Unbeknownst to me that person used an artificial intelligence application to create the brief and the cases included in it were what has often being (sic) described as ‘artificial intelligence hallucinations,’” he wrote.”It was absolutely not my intention to mislead the court or to waste respondent’s counsel’s time researching fictitious precedent.”
The apology was good, Odenwald wrote, but insufficient to correct the errors.
It’s worth taking a look at what Karlen actually wrote:
Second, I would like to apologize to both The Court and respondent's counsel for my filing on May 1, 2023. I was unaware of any of the fictitious cases presented until I received respondent's motion to strike on June 23, 2023. My inability to afford counsel led me to hire a consultant from an online marketplace representing themselves as a licensed attorney in California offering to prepare the brief for a fee less than one percent of the cost of the retainers sought by attorneys to actually represent me. Unbeknownst to me that person used an artificial intelligence application to create the brief and the cases included in it were what has often being described as "artificial intelligence hallucinations". It was absolutely not my intention to mislead The Court or to waste respondent's counsel's time researching fictitious precedent.
Marc Lauritsen had the best take on this, in my opinion:
Dear Unrepresented Litigant: We know you can't afford a lawyer, can't get legal aid, and neither our clerks nor our judges are allowed to give you legal advice. But if you file something that includes a made-up case you got online, we're going to bust you.
The fine for this guy was $10,000. In my view, keeping in mind I’m not a smart person, the purposes of court sanctions should be to deter bad conduct by attorneys / litigants and to show others that there are real consequences for things like missing deadlines and misleading the court. What’s the takeaway here for Karlen and other unrepresented litigants? Hire an attorney that you can’t afford? Pay the guild or get fined by the guild?
The Year Chatbots Were Tamed
This article by Kevin Roose talks about how the AI models by Microsoft and other big tech companies have become more and more uninteresting since Microsoft’s Sydney tried to convince him to leave his wife:
Despite all the progress being made in artificial intelligence, today’s chatbots aren’t going rogue and seducing users en masse. They aren’t generating novel bioweapons, conducting large-scale cyberattacks or causing any of the other doomsday scenarios envisioned by A.I. pessimists.
But they also aren’t very fun conversationalists, or the kinds of creative, charismatic A.I. assistants that tech optimists were hoping for — the ones who could help us make scientific breakthroughs, produce dazzling works of art or just entertain us.
We’re probably seeing a leveling out of the models that the big tech companies are offering to the public. If we presume that people are going to ChatGPT / Bard / Claude for legal advice, do we want the tech companies deciding that they can’t get it? Do we want a free-for-all? Or do we want something better? Right now it seems like we’re letting the tech companies make that decision for us.