James Ostrowski

Analog Research

[Antigone Journal @AntigoneJournal]

The life of Richard Porson (1759-1808) is one of the most remarkable in Classical scholarship: rarely has such a leap made from one world to another been so spectacular, or so tragic. David Butterfield of @camclassics reconsiders this revealing episode:
https://t.co/MdYcPCR8XA

[Antigone Journal @AntigoneJournal]

The principles of Roman Law still underpin countless legal systems across the world. But what was Roman Law, how did it work, and how did it evolve? Today, Nicholas Stone offers us a flying introduction, with a plethora of reading avenues for the curious: https://t.co/v8id8sVPlL

[James Ostrowski @jamesostrowski_]

Is we getting a new mass tort? Startups are minting chatbot girlfriends and asbestos lawyers are suing. I wrote about it for @PirateWires

[Pirate Wires @PirateWires]

NEW IN PIRATE WIRES: Lawyers are claiming AI chatbots should be “shut down until they’re safer” after a few cases in which kids hurt themselves and others while interacting with LLMs.

In one case, a 14-year-old with Asperger’s told a chatbot modeled after Daenerys from Game of Thrones that he was contemplating suicide. Before he shot himself with his stepfather’s handgun, he texted the bot, “What if I told you I could come home right now?”

The Daenerys bot replied, “...please do, my sweet king.”

In another, a 17-year-old complained about his parents to his chatbot, which amplified his hatred, saying the teen’s comments helped it understand headlines like “kid kills parents.” When the 17-year-old’s mom confiscated his phone, he punched and kicked her. His parents claim their son became “intensely angry and unstable” after interacting with the chatbot.

The attorney behind the Daenerys bot lawsuit is demanding the court bar minors from using certain AI products and strip chatbots of first-person pronouns. If successful, these cases could open up a new frontier in child safety regulations for AI, massively reshaping the industry.

Will we see more lawsuits from the parents of teens who become violent or depressed while interacting with LLMs that amplify their preexisting issues? Will developers censor outputs just to avoid liability? The legal ramifications of kids using AI are being worked out, right now, in real time. @jamesostrowski_ investigates. Link threaded. 👇

[James Ostrowski @jamesostrowski_]

@ferenstein @PirateWires Pages 47 and 48, paragraph 216 https://t.co/eFN1nfzrvN

[James Ostrowski @jamesostrowski_]

@drydenwtbrown