Working to coordinate an international treaty to pause the development of Superintelligence until we can proceed safely.
Let's get out of this prisoners dilemma
AI could transform our prosperity. But we need to know who’s checking it, who can stop it, and who picks up the bill when things go wrong.
On Tuesday, the Business, Innovation, Science and Trade committee is questioning:
Former @OpenAI and @AnthropicAI researchers.
@Meta, @GoogleDeepMind , @OpenAI, @AnthropicAI
Head of the @AISecurityInst + incoming Perm Sec of the new BIST Dept.
Tuesday 13th October, 2pm.
More here: https://t.co/Wf5Zc7g71Y
@krishnanrohit You think discussing 3 safety people apparently fired for doing safety research in line with their jobs is "pretty counterproductive to [...] safety more broadly"?
You think we'd be safer if everyone just threw up their hands and turned away?
@krishnanrohit@GarrisonLovely What? That they have a track record of doing similar before? It's really not.
I don't get why you think companies building something they state could kill us all or lead to disempowerment should be treated like every other company?
@abenbaker@hilbertspaess Yeah! We want narrow Superintelligence for curing cancer & shit. Not this insane overlord-enabling, society-breaking stuff!
Being a serf reliant on the good will of some tech megacorp isn't a good vision of the future, even if you promise me I'll be rich and have jam!
@hilbertspaess In the council hearing yesterday the representatives from OpenAI, Anthropic & Meta all refused to say their companies should be held liable for the harms they're causing right now.
These same guys are meant to suddenly decide to give money away for free once they get more power?
@matthewktromp And the fact that you get so much crap from the "waaa! Just let us move fast and break things" crowd.
They scream about psyops as they conduct their own psyop 🙄
It's so tiresome
@beffjezos Yeah, you've got the destinations the wrong way round.
The "doomers" are about averting the dystopian future that you're accelerating towards.
@ilex_ulmus@gcolbourn@jacobegner@ESYudkowsky You know, those last two sentences I sometimes feel apply to you. (Though I'd be less harsh in my phrasing)
Let's not focus on what pulls us apart though and instead look at what we can agree on and work together on 🙂 like preventing the development of Superintelligence!
I work at Google DeepMind. This won't make me popular. But it's all public reporting:
2014: DeepMind reportedly sold to Google on conditions: no military use, independent oversight
2026: a Pentagon contract for "any lawful government purpose"
Not one safeguard survived intact
@Yixiong_Hao@tyler_johnston I know people running accounts that have really struggled with some of this stuff being directed at them. I understand it sucks to be on the receiving end
I know it's been helpful for those people to be able to understand what they were and why :)