"We did it, Patrick! We made a technological breakthrough!"
A place for all those who loathe AI to discuss things, post articles, and ridicule the AI hype. Proud supporter of working people. And proud booer of SXSW 2024.
AI, in this case, refers to LLMs, GPT technology, and anything listed as "AI" meant to increase market valuations.
Seems like this should be illegal somehow... Though I'm not sure what it would actually fall under.
Possibly libel? If he found any of the slop to be low quality or objectionable, he could say it's disparaging against him (and knowingly false) for the company to claim that he wrote it.
fraud, defamation, appropriation of name or likeness, fraudulent misrepresentation...
Yeah I can't just write a book put Brandon Sanderson's name on the front and collect the proceeds.
But what if ai writes a book and puts Brandon Sanderson's name on it because it was trained on his books and was told to write a book that looks like one of his, because you fired him, and then you don't proof-read the result and publish it with Brandon Sanderson's name on it?
Speaking hypothetically asking for a friend.
I mean Amazon is a mecca of self publishing
Maybe Fraud?!
It would be fraud when they were presenting it to other people, but I'd be curious if it would apply to him, since he's not the one being defrauded.
Is impersonation by misattribution actionable in a court of law?
Assuming here the USA (because it's mostly the USA that does this kind of rabid shit): most states have "right of publicity" laws in which individual have exclusive right to control the commercial use of their name, likeness, or other recognizable aspects of their persona. Commercial use of someone's such identity and/or persona can be very, very severe depending on which state the offending action was taken in.
More federally we have the Lanham act that explicitly prohibits the use of a person's name/identity in commerce in a way that is likely to cause confusion about affiliation, sponsorship, or endorsement. This doesn't even require a trademark registration, though it is easier to get your Lanham rights if you have one.
Finally there's the hallucination angle which can lead to defamation suits. The LLM will inevitably hallucinate, and if the hallucination is of the variety that could expose the named person to ridicule, hatred, etc. that is a straightforward defamation suit that could result in a whole lot of monetary loss to the publisher.
And here's the fun part. These are not exclusive. A publisher could face statutory damages for right of publicity violation, stripping of profits for the Lanham violations, and compensatory and punitive damages for defamation by hallucination.
Note: IANAL, obviously, but I do research things for a living. Consult with an attorney, not me, for actual court cases.
All LLMs do is hallucinate...
Bad assumption, it turns out, and the article tells us which EU regulation helped him.
(It's at the bottom, since we're lazy)
Wait, a EUROPEAN company has gone slop-happy and anti-human evil?
Fuck.
The American rot is spreading quickly, isn't it?
The American rot
This may surprise you but capitalism thoroughly predates the US.
Not all capitalism is quite as out of control and flat-out evil as American capitalism. Some countries still try, at least, to keep their capitalists under control (with varying degrees of success, natch). The USA seems to think it exists for the capitalists.
Europe had half a dozen east India companies fighting over superexploitation of the east, with armies bigger than the national armies. It's just regular capitalism.
In today's lesson you learn about verb tenses? 'Cause you need it.
Are you fucking stupid? You inherited our rot, practice it in the same way, and claim you're special and unique for it. You're not. America isn't exceptional, its just European.
I am no lawyer but it seems like a slam dunk lawsuit, especially since they are an ex-employee. If it was just some other person or journalist's name being used they could at least make the claim that they made the name up and it's just a coincidence and were just trying producing your garden variety AI slop that so many other news outlets are.
Trademark law seems like it'd be the easiest. Pretty sure that a journalist's byline would count as a brand, and posting AI-slop under that brand without permission seems like it would tarnish that brand's reputation.
It could work, but only if he registered his name as a trademark. Trademark protection isn't automatic -- it's only enforceable if you've registered the trademark before the alleged infringement happened.
Depends on the contract he chose to sign when taking up employment with them.
Most employment contracts are illegal. They throw in bad terms just in the hope an employee won't fight.
Things should clear up soon once sufficient legal advice becomes trivially cheap.
Hold on, let me get my AI lawyer.
Litigation is probably still going to need a person but working though jargon on contracts you wouldn't normally think to call a lawyer for to begin with will become trivial.
They don't go to litigation because they know its illegal. Most common example are non-compete clauses. No one can stop you from gainful employment wherever the fuck you please. The courts have squashed this B.S. thoroughly, except on extremely narrow circumstances, and no one wants the added costs. Litogate, sure lose and pay both costs and still lose.
Everyone had “upscaled” their productivity with AI. Then he was told he was getting let go because Esports Insider had been de-indexed by Google.
Mmmhmmm. Company lessons learned, right?
Nope.
It was a "slop in the face."
FTFY
Felt like a Futurism headline, but on yahoo domain.
The image credits futurism.
What happened here?
yahoo and some other sites like msn republish articles from other sites for free, hopefully with a valid license.
Hopefully. If Yahoo were also republishing other sites articles without permission, that would be rather ironic, given how similar that feels to the topic in question.
Here's the proper link: https://futurism.com/artificial-intelligence/fired-journalist-ai-posts-in-his-name
Being fired was only the first indignity that one freelance writer had to suffer.
Even though Ben Touati was no longer working at ClickOut Media, new articles continued to appear under his name. An AI, it turned out, had taken over his byline, churning out stuff that was far below the standards he held for his real work.
The situation felt like a “slap in the face,” Touati told Press Gazette. The articles started appearing just days after he was fired from ClickOut’s German operation.
“The five articles were just like lazy, obviously slop, obviously there’s not a real person that is behind that,” the Stockholm-based writer fumed.
A statement from ClickOut Media didn’t address why it was effectively impersonating a former employee with AI.
“We use AI-assisted content where appropriate in tandem with human checks and edits,” it said in the statement, per Press Gazette. “We continue to evolve our AI agents to be more accurate and improve our human editorial processes.”
Heaps of online publishers are embracing AI — often deceptively — and Clickout Media is right up there with the worst of them.
Scumfucks. What I hate about these sites having to resort to all sorts of trickery to keep being relevant and having more traffic -- and ad revenue -- coming in. Won't be surprised either some of these sites get paid by other corporations and/or governments to push specific narratives and propaganda.
I thought this was the onion for sure with that title.