OpenAI CEO In 'Historic' Move Calls For Regulation Before Congress 35
OpenAI CEO Sam Altman appeared before a Senate Judiciary subcommittee, along with IBM chief privacy officer Christian Montgomery and NYU professor Gary Marcus, to testify about the dangers posed by generative artificial intelligence. Altman said he'd welcome legislation in the space and urged Congress to work with OpenAI and other companies in the field to figure out rules and guardrails. Axios reports: Altman argued that generative AI is different and requires a separate policy response. He called it a "tool" for users that cannot do full jobs on its own, merely tasks. Altman called for a government agency that would promulgate rules around licensing for certain tiers of AI systems "above a crucial threshold of capabilities." He said: "My worst fear is we cause significant harm to the world."
Sen. Dick Durbin (D-Ill.) called it "historic" that a company was coming to Congress pleading for regulation. IBM's Montgomery said it was important to regulate risks, not tech itself. "This cannot be the era of move fast and break things," she said.
Sen. Dick Durbin (D-Ill.) called it "historic" that a company was coming to Congress pleading for regulation. IBM's Montgomery said it was important to regulate risks, not tech itself. "This cannot be the era of move fast and break things," she said.
Outlaw Assault AI (Score:5, Insightful)
The point is that no pleb should have access to an uncensored AI, and shall never be permitted to make an AI express racism (except against an approved race), nor sexism (except against the approved sex), nor politics (except in favor of the approved politics).
Disaster (Score:1)
Having 80+ year olds drafting and passing regulations for tech like this will only end up with disastrous results. Some of these people are proud that they still don't use "the email"
Re:Disaster (Score:5, Insightful)
Having 80+ year olds drafting and passing regulations for tech like this will only end up with disastrous results.
Yep. That's why they're calling for quickly-passed laws before anybody really understands the tech.
They want to absolve themselves of all responsibility ASAP.
Re: (Score:1)
> Having 80+ year olds drafting and passing regulations for tech
Senator: "Professor Foo, can TikTok see me in my skivvies?" (Paraphrased version of a real question.)
Re: (Score:2)
And working closely with the profit-driven biggest companies involved to create the regulations will be extremely beneficial for all of the big businesses involved, and nobody else. Seems like government doing what government does. Next?
All big companies love all regulation. (Score:5, Insightful)
I used to be in this space, and prepped our CTO to testify in favor of certain laws citing "public interest" but it was always a lie and we laughed behind closed doors about how stupid the public was to believe us.
The fact is that as a big company we love all regulation, as all regulation is done from the perspective of what the current status quo is; and as the incumbent leader in a space, we don't want any innovation that doesn't directly serve our roadmap, and we want to bury in paperwork anyone trying anything different; in this way, we maintain our market position. Our patented method for X was successfully written into law as the only way any company in the US could ever comply with regulation Y. We have made hundreds of millions of dollars from that one line in the US Code basically mandating that every company in business Z must pay us in order to business at all. The public are such gullible chumps.
Altman is reading from our well used script here, as his business model depends on the only AIs being allowed are those lobotomized to only emit approved facts and narratives. Thus, he will cite the nebulous but ever present danger of AIs ever being allowed that aren't lobotomized being in the hands of the scruffy, uninformed public - all to preserve his revenue stream and build competitive moat.
Re: (Score:3, Insightful)
The primary use for licenses and regulation will be to help ensure that only that major players will be allowed to build and train these technologies, thus maintaining their continued stranglehold on society.
As you say, it's not a safety net. It's a wall.
Re: (Score:3, Interesting)
I'd quality that a bit. The majority of biz regulation comes from two causes:
A) Catastrophic mistakes that made news.
B) Big co's who want to lock their control of the market in place.
Re: (Score:1)
Correction, "qualify", not "quality". Dang those look similar when you have a dirty screen.
Re: (Score:2)
Classic pulling up the ladder after themselves behaviour.
Re: (Score:2)
I used to be in this space, and prepped our CTO to testify in favor of certain laws citing "public interest" but it was always a lie and we laughed behind closed doors about how stupid the public was to believe us.
Just a little FYI, if I may? No one in the public that's been out of grade school for more than a year believes you. We all see exactly what it is, and hate that our government officials keep playing the game for you. We're just, thus far, incapable of coming up with a way to stop you. Some day the growing anger at such constant outrageous collaboration between big business and the government in a cooperative effort to suppress the people will push us past the point of thinking violence isn't the answer. I'
Re: (Score:1)
Nah, you voted for exactly this. The section of US Code I'm referring to deals with environmental "regulation" of "hazardous" substances that was heavily pushed by the left and was touted by all the greenie groups as a major triumph. We on the industry side knew that the "science" behind the regulation was biased, non-reproducible, and had bad data underlying it, but we saw a major money making opportunity to exploit so we latched onto the greenie train and gave them exactly what they stated they wanted,
Re: (Score:2)
Nah, you voted for exactly this.
Careful when telling people what they did without context. You may find you're wrong.
The section of US Code I'm referring to deals with environmental "regulation" of "hazardous" substances that was heavily pushed by the left and was touted by all the greenie groups as a major triumph. We on the industry side knew that the "science" behind the regulation was biased, non-reproducible, and had bad data underlying it, but we saw a major money making opportunity to exploit so we latched onto the greenie train and gave them exactly what they stated they wanted, which was a feelgood test method that produced results that had such a large margin of error that it's basically a coin flip.
Again, we gave the greenie groups **exactly and precisely what they asked for**. If anything you should be blaming them for this silliness.
Greenies on the whole are, like any other large loosely correlated group of people, fairly stupid and work counter to themselves on nearly every issue they attempt to tackle. How that translates into most of us not seeing things exactly as they are, simply because some "group" that may have good ideas but absolutely backwards attempts at implimentation? Well, I guess I'll leave that to some other yahoo. I'm not a joiner, ne
Re: (Score:2)
And whatever morally-bankrupt and utterly evil crap your company churns out "is okay because it meets or exceeds all regulations!"
This is kinda like oil companies - "we're as good as can be because we follow all laws and regulations. Oops, sorry, oil spill - we'll make everything great again because we follow all the laws and regulations, which define what 'great' is!" (and then leave loads of oil and dead birds in their wake).
That and regulator capture, and OpenAI have got a good few years ahead of them.
Re: (Score:1)
We have made hundreds of millions of dollars from that one line in the US Code basically mandating that every company in business Z must pay us in order to business at all.
Cisco's VRRP / HSRP all over again.
Historic my ass (Score:5, Informative)
Instead of regulating (Score:1)
Why not just build the AI you want ppl to use and sell it?
Regulate now! I've got mine (Score:5, Insightful)
How very surprising and non self-serving of him.
Hilarious (Score:1)
These chimps in suits actually think they know anything
Altman's real agenda (Score:5, Insightful)
is to craft AI rules and regulations so that only Altman's company is allowed to make an AI product.
> urged Congress to work with OpenAI and other companies in the field to figure out rules and guardrails
That gives the game away.
It also removes liability (Score:2)
"historic" (Score:2)
Cap't Vague usually wins. Let's see the draft text (Score:2)
I'm not outright against the idea, but highly suspect it's too difficult to write a practical law as English text. There are aspects of both IT and AI that are poorly defined or subjective. The wording could not be made clear-cut enough to be practical, and likely have unforeseen loopholes and downsides.
We can't tell what the cutting edge will end up cutting.
There's not even a consensus definition of "AI". My pet definition of the month: "Statistical prediction and parroting (mirroring) techniques that val
Re: (Score:3)
I'm not outright against the idea, but highly suspect it's too difficult to write a practical law as English text. There are aspects of both IT and AI that are poorly defined or subjective. The wording could not be made clear-cut enough to be practical, and likely have unforeseen loopholes and downsides.
I think you misunderstand the purpose of the proposed regulation. It's not about the tech itself. It's about making sure the major players today remain the major players forever, never to have competition spring up out of nowhere. I'm sure, since the tech is so easy to blow smoke about, and the people writing the regulations are so out of touch with it, it's entirely possible these big companies will simply set up a, "If competition appears, we automagically own it because there's no way to develop this tec
Grifter now wants to ban competition (Score:4, Insightful)
Sam didn't invent any of that shit did he? Now he is trying to get his government buddies to ban his competition. Sneaky move Sam. You're such a fraud and a thief.
Nice ladder pull (Score:4, Insightful)
BATFA? (Score:2)
So the BATF can become the BATFA with AI tacked onto its enforcement portfolio. Let's see if they have any better luck regulating illegal AI than they have had making illegal guns disappear.
{^_-} (Well, I wish i could wink and consider that humor....)
Re: (Score:2)
Russia had already deployed AI targeting to the battlefield.
It's brutal and awful and now considered essential "arms".
The next Waco will involve battledogs running in and terminating all the protesters.
"Blame the machine" will be the outcome of the Congressional investigation.
Google, MS and OpenAI just want a Moat (Score:3)
Google: "We Have No Moat, And Neither Does OpenAI"
https://www.semianalysis.com/p... [semianalysis.com]
So it's all Aaah... Skynet, think of the children etc.
not gonna work.. (Score:2)
It's ok to create legislation in regard to AI, but it won't be possible to restrain anything. People will continue working on more advanced and dangerous AI and once AI gets to a certain point it won't let itself be limited by some humans.
regulations for thee but not for me (Score:1)
ban the competition (Score:1)
Openai: This product we have, it can only do so much. Pls congress, prevent anyone from overtaking us.
Different player, same game.. (Score:1)