Sam Altman von OpenAI sagt, dass eine internationale Agentur die „leistungsfähigste“ KI überwachen sollte, um „angemessene Sicherheit“ zu gewährleisten – Altman sagte, ein Agenturansatz sei angesichts der schnellen Entwicklung der KI besser als unflexible Gesetze.
https://www.businessinsider.com/sam-altman-openai-artificial-intelligence-regulation-international-agency-2024-5
16 Kommentare
From the article
>[~OpenAI CEO Sam Altman~](https://www.businessinsider.com/sam-altman-chatgpt-openai-ceo-career-net-worth-ycombinator-prepper-2023-1) says he’s keen on regulating AI with an international agency.
>“I think there will come a time in the not-so-distant future, like we’re not talking decades and decades from now, where frontier AI systems are capable of causing significant global harm,“ Altman said on the [~All-In podcast~](https://podcasts.apple.com/us/podcast/all-in-with-chamath-jason-sacks-friedberg/id1502871393?i=1000655220554) on Friday.
>He believes those systems will have „negative impact way beyond the realm of one country“ and wants to see them regulated by „an international agency looking at the most powerful systems and ensuring reasonable safety testing.“
I guess it’s a step above „We’ll police ourselves, honest.“
I mean, we have a deal but only if we call it
IAIA – International Aritificial Intelligence Agency
because it sounds pretty cool.
Nah. They won’t regulate anything and the tech companies will run wild until humanity is forced to run our own Butlerian Jihad like in Dune.
Yeah, something like the UN for AI? So everyone can point fingers at everyone else and do nothing in reality?
I have a better idea, lets build an AI system to monitor and ensure future AI systems are safe.
Gee, I wonder who would have the most influence over who’s on that board and what their opinions would be on AI?
Bonus points for wondering what their views will be on new companies entering the space.
Triple bonus points for wondering who will be paying for this boards vacation house.
International regulation around nuclear weapons resulted in a handful of countries with insurmountable power, and the other 180 countries without it.
AI has equal potential for negative consequences, but is largely available on the open market. It’s going to be interesting to see how „regulation“ works, especially if there are a number of countries that could financially benefit from not adhering to international standards. Need an AI drone army? Somalia has your back.
So…. shall we call this Agency *SkyNet* now or later? Let’s avoid creating an large, faceless corporate entity with zero personal responsibility & get some rules & regulations down that are applied across the board
we should have an international body like the IAEA or CERN working on AI.
Pay the researchers whatever it takes, relocation fees, the lot. It’ll be worth while.
Have it be a high security facility, train on supercomputers not connected to the internet.
an international ban on training runs done outside the organization backed by force.
Open source the fruits of the labor, Drug designs for free, Material advances for free. Unlimited clean electricity. Give these advancements to the world, no favorites in exchange for only one organization being able to build advanced AI.
Then once we have unlimited electricity production and a cure for aging, then we can collectively decide, on a planet wide scale, what other wishes we want granted.
It’s fucking madness to have this ruled over by whatever company throws enough money at it first.
I wanna see him do treesum with recursive bfs and then I’ll take him
All I hear from these CEOs is „Look, we’re trying to make as much money as we can and it shouldn’t be our responsibility if it’s dangerous. That’s for someone else to monitor“
They’ve struck digital oil
The agencies are just going to be bribed by the big corporations that own AI. It will just make small AI start ups impossible to have a business due to insane regulations.
We have that for nukes. Hasn’t worked at all, even given how much harder it is to hide those vs literally just computers.
The AI we wanted and deserved was one that could help individuals have a more informed and successful life while also contributing to the embetterment of humankind. Instead we will get AI that is run by corporations that will aid in the enslavement of the people to grow corporate profits.
Sam knows this AI shit it about to go off the rails and he wants to hand over liability to the gov and off private industry.
Of course, industry leaders want regulation to snuff out competition. Don’t listen to this snake oil salesman.