Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

FTC investigating ChatGPT over potential consumer harm

npr.org

Post reply on HN
Print view View original

21–30 of 146 posts

Re: FTC investigating ChatGPT over potential consumer harm

#21

Post by hackernewds »

biofunsf wrote: post #16
andrewstuart wrote: post #7

Sam Altman himself sat in front of the top levels of government and freaked them out by telling them his company was putting in place the groundwork to possibly end humanity. A really strange thing to do, I remain puzzled as to why any CEO would stir up the government like this. Having panicked regulators with science fiction, I hope he's not surprised when they take action.

Maybe he's just a CEO doing the ethically responsible thing and not the purely self-serving thing for the sake of shareholders? (though as humans, the shareholders have some stake in humanity too)

The fact that we are bewildered instead of cherishing, and punishing instead of rewarding shows us what incentives lie in honesty nowadays. Sounds nihilistic but Sam could have done better for his company, by being untruthful

Re: FTC investigating ChatGPT over potential consumer harm

#22

Post by biofunsf »

silisili wrote: post #4
nonethewiser wrote: post #2

> The agency says it's looking into whether the AI tool has harmed people by generating incorrect information about them Havent heard this complaint before. And I wouldn’t really call this harmful to consumers. If anything its harmful to the subject, but consumers in general. But even so it doesnt meet the definition of libel. How harmful is it really? Of all the complaints against OpenAI, absolutely none of them see…

It could absolutely cause real harm to any one person or group at a time. There was a recent case where it accused a professor of sexual harassment, seemingly out of nowhere. Imagine it spreading on Twitter and getting mobbed, doxxed, etc. Or a potential employer seeing it and silently rejecting your employment. Excuse the link, the big papers have awful popovers, this one seemed acceptable... https://decrypt.co/1257…

At the bottom of the ChatGPT prompt it always says: "ChatGPT may produce inaccurate information about people, places, or facts."

So... is this just a case of user error? Or maybe the FTC needs to require a larger font size or just better disclosure? Mostly joking, but not entirely since I do think it's too small. Here's a proposal: https://i.imgur.com/IGON20r.png

Re: FTC investigating ChatGPT over potential consumer harm

#23

Post by belval »

biofunsf wrote: post #16
andrewstuart wrote: post #7

Sam Altman himself sat in front of the top levels of government and freaked them out by telling them his company was putting in place the groundwork to possibly end humanity. A really strange thing to do, I remain puzzled as to why any CEO would stir up the government like this. Having panicked regulators with science fiction, I hope he's not surprised when they take action.

Maybe he's just a CEO doing the ethically responsible thing and not the purely self-serving thing for the sake of shareholders? (though as humans, the shareholders have some stake in humanity too)

Wow that's naïve (or sarcasm).

I did an internship with ElementAI a while back. They had a full lobbying division to "come up with a legislative framework with the government for ethical AI". This is quite literally just a business strategy where you get to write the rules because lawmakers see you as "their guys" while challengers are seen as reckless.

It was already a thing in 2019 and it definitely is still a thing. Sam Altman just made the gambit that congress would see him as the honest and prudent guy. Seems like that just didn't work.

Re: FTC investigating ChatGPT over potential consumer harm

#24

Post by bodecker »

biofunsf wrote: post #18
arthurcolle wrote: post #17

Earlier quoted context omitted.

Yep. Especially with geohot leaks and torrent activity at an all time high - OpenAI looks cooked

What are the geohot leaks?

I assume comments like these, "GPT-4: 8 x 220B experts trained with different data/task distributions and 16-iter inference."

https://twitter.com/soumithchintala/status/16712671501017210... https://archive.li/rfFlW

I'm not sure the most canonical paper on mixture of experts but here's one possible:

https://arxiv.org/pdf/1701.06538.pdf

Re: FTC investigating ChatGPT over potential consumer harm

#25

Post by ZunarJ5 »

biofunsf wrote: post #18
arthurcolle wrote: post #17

Earlier quoted context omitted.

Yep. Especially with geohot leaks and torrent activity at an all time high - OpenAI looks cooked

What are the geohot leaks?

https://the-decoder.com/gpt-4-architecture-datasets-costs-an...

Not op, but this is where a cheeky google got me.

Re: FTC investigating ChatGPT over potential consumer harm

#26

Post by justinclift »

silisili wrote: post #4
nonethewiser wrote: post #2

> The agency says it's looking into whether the AI tool has harmed people by generating incorrect information about them Havent heard this complaint before. And I wouldn’t really call this harmful to consumers. If anything its harmful to the subject, but consumers in general. But even so it doesnt meet the definition of libel. How harmful is it really? Of all the complaints against OpenAI, absolutely none of them see…

It could absolutely cause real harm to any one person or group at a time. There was a recent case where it accused a professor of sexual harassment, seemingly out of nowhere. Imagine it spreading on Twitter and getting mobbed, doxxed, etc. Or a potential employer seeing it and silently rejecting your employment. Excuse the link, the big papers have awful popovers, this one seemed acceptable... https://decrypt.co/1257…

Real world example of how things can go very, very badly: :(

https://en.wikipedia.org/wiki/Indian_WhatsApp_lynchings

Re: FTC investigating ChatGPT over potential consumer harm

#27

Post by janalsncm »

Well that’s one way to solve AI alignment. Just sue the companies for harming consumers.

I’m half joking, but in the statistical distribution of good, bad and ugly things automation can do, we will need to draw some line as to what is legally actionable. For example, if I Google search recipes for explosives, Google isn’t liable for surfacing accurate information. And imo OpenAI isn’t liable if their chatbot gives the same accurate info in response to the same query in the form of a chat message rather than a link.

Re: FTC investigating ChatGPT over potential consumer harm

#28

Post by biofunsf »

belval wrote: post #23
biofunsf wrote: post #16

Earlier quoted context omitted.

Maybe he's just a CEO doing the ethically responsible thing and not the purely self-serving thing for the sake of shareholders? (though as humans, the shareholders have some stake in humanity too)

Wow that's naïve (or sarcasm). I did an internship with ElementAI a while back. They had a full lobbying division to "come up with a legislative framework with the government for ethical AI". This is quite literally just a business strategy where you get to write the rules because lawmakers see you as "their guys" while challengers are seen as reckless. It was already a thing in 2019 and it definitely is still a thin…

Not sarcasm! Though I agree this also helps build Open AI's moat against competitors... but that doesn't make Altman's statement not true. It's still self-serving, but I think if he was totally willing to lie he'd have made other self-serving statements, and not this particularly high-risk, shocking, and also true (IMHO) self-serving statement.

Re: FTC investigating ChatGPT over potential consumer harm

#29

Post by 1vuio0pswjnm7 »

ZunarJ5 wrote: post #25
biofunsf wrote: post #18

Earlier quoted context omitted.

What are the geohot leaks?

https://the-decoder.com/gpt-4-architecture-datasets-costs-an... Not op, but this is where a cheeky google got me.

"The idea is nearly 30 years old and has been used for large language models before, such as Google's Switch Transformer."

Innovation! :)

Re: FTC investigating ChatGPT over potential consumer harm

#30

Post by ta988 »

biofunsf wrote: post #18
arthurcolle wrote: post #17

Earlier quoted context omitted.

Yep. Especially with geohot leaks and torrent activity at an all time high - OpenAI looks cooked

What are the geohot leaks?

George Hotz (pseudo geohot) in his recent interview with Lex Fridman gave some info on the probable structure of gpt4.
Post reply on HN