wobwegerJun 21, 2023
Timnit Gebru (she/her).

"“OpenAI repeatedly argued to European officials that the forthcoming AI Act should not consider its general purpose AI systems…to be “high risk,” a designation that would subject them to stringent legal requirements including transparency, traceability, and human oversight."

Oohh noooz our stuff is so strong it might render humanity extinct.

But also no no our stuff shouldn’t be classified “high risk.”

By @perrigo

https://time.com/6288245/openai-eu-lobbying-ai-act/

Exclusive: OpenAI Lobbied the E.U. to Water Down AI Regulation

OpenAI lobbied the E.U. to weaken forthcoming AI regulation, even as the company calls in public for stronger AI guardrails, documents show.

Time
Jun 21, 2023 at 2:25pmWeb
Show threadTimnit Gebru (she/her).Jun 21, 2023
Why didn't they just publicly release the white paper if they stand by everything they're doing? 🤔
Show threadBen WaberJun 21, 2023
@timnitGebru the only thing accurate about their name is that they're "artificial"
Show threadGaëtan PerraultJun 21, 2023

@timnitGebru @perrigo

"Don't use my public statements about the product, please only use the ones we make in private. Behind closed doors.

"This is how things work in America."

Show threadCorbinJun 21, 2023
@timnitGebru @perrigo It's so transparent. In the USA, they're begging for regulation from weak regulators who are divided and partisan; in the EU, they're trying to dodge regulation from a competent federation of dozens of jurisdictions.
Show threadBernd PorrJun 21, 2023
@timnitGebru @perrigo but, but the extinction only happens in 1000s of years and distract from the current problems. Obv nobody at the #EU would see through that bullshit. I'm glad they did and stick to it.
Show threadBen CarverJun 21, 2023
@timnitGebru @perrigo very, very well put!
Show threadMorten Juhl-JohansenJun 21, 2023
@timnitGebru @perrigo Either this is A GAME-CHANGER PARADIGM, or this is a toy. I feel like it should not be accepted to market both views at once...
Show threadKona GoodhartJun 21, 2023
@timnitGebru @perrigo
Show threadJon (now at neuromatch.social)Jun 21, 2023

@timnitGebru who could have predicted?

great reporting, @perrigo !

Show threadJill FellowsJun 21, 2023

@timnitGebru @perrigo

"Look, it's high risk. Everyone pay attention and give us free publicity!!"

"No, not like that, Europe!"

Show threadFuzzysteveJun 21, 2023
@timnitGebru @perrigo "our stuff shouldn't be restricted. our competitors should be." Am I just being cynical?
Show threadLinebyline→mastoart.socialJun 21, 2023
@timnitGebru @perrigo Sounds like the wackyland upside-down version of that thing fascist propagandists do where the enemy is an ever-present existential threat but also pathetically weak and inferior.
Show threadTreesJun 23, 2023
@timnitGebru @perrigo Meanwhile chatGPT continues to interfere w/student's ability to complete their course work so they can graduate by claiming their papers were not written by students but by it. And lawyers are continuing to deal with getting disbarred by relying on it for case law even though chatGPT just makes up case law while claiming its legit, also while it's fabricating lists of published science papers that got scientists named in fabrications filing defamation lawsuits! Fuck OpenAI!