R. Donald J. Gauvreau

Comms @MIRIBerkeley. Opinions mine, not MIRI's. ᄽ●・●ᄿ

[R. Donald J. Gauvreau @cmSalticidae]

Fukuyama: "The End of History was about the triumph of liberal democracy."

Frontier labs: "At long last, we have created the end of history."

[Francis Fukuyama @FukuyamaFrancis]

We Need an International Treaty to Ban Superintelligence https://t.co/39XHsrlqxC

[Harlan Stewart @HumanHarlan]

On Feb 29, 2020, I was at a wedding with relatives and family friends. I ended every handshake by using sanitizer and explaining that, although there were only 22 confirmed US cases, a pandemic seemed likely. This might be the last big gathering we would attend for a while, I said. And we might need to cancel upcoming travel plans.

This wasn't actually very far-sighted of me. It was only 11 days before the WHO declared a pandemic, and the signs were abundant. But it was still early enough that what I was doing was very obviously strange. I could feel the awkwardness.

I think I'm unusually comfortable with that type of awkwardness. I've also had a lot of practice with it.

In 2014, I read the argument that superhuman AI might make humanity extinct, found it to be convincing, and began trying to explain it to the people around me. I could feel the awkwardness.

AI risk is easier to discuss today than in 2014. The signs are more abundant. But there's still a sort of collective awkwardness about it. A reluctance to look at what's currently unfolding, extrapolate where it will take us in the coming years, and then *actually do something about it.*

I think individuals have an opportunity to create a lot of good by breaking through that awkwardness.

I invite you to be the weird guy with hand sanitizer at the wedding.

[Nate Soares ⏹️ @So8res]

Wikipedia spends more effort convincing you to donate $2 than AI companies spend galvanizing the world to stop the race. There's a real inconsistency between what they say and how they behave. This mealy-mouthed behavior incurs real costs. I hope AI folk learn to recognize that.

[R. Donald J. Gauvreau @cmSalticidae]

@mh012012 @AndyMasley Anthropic could require 2 of 3 of [birth certificate, passport, SSN] to verify a customer is a US citizen, but could still be liable if someone fooled them, & now they'd have to protect a lot of otherwise useless PII.

and a foreigner could still log into a verified-US account...

[Nate Soares ⏹️ @So8res]

"The government will never do anything to hinder AI," they said.

Inaction yesterday does not imply inaction today. An export control directive came out of nowhere, and a ban on superintelligence could too. Inevitabilism is wrong.

[Nate Soares ⏹️ @So8res]

Leadership at Google DM, Anthropic, OpenAI, and xAI have all said they'd prefer the whole world slow down AI. There's still a ways to go — wikipedia puts more effort into convincing you to give $2 than AIcos put into alerting the world to the danger of ASI — but it's a start.

[BBC News (UK) @BBCNews]

Version of AI tool too powerful for public released to public https://t.co/2HtI4UJCTW

[Nate Soares ⏹️ @So8res]

I've got a lot of quibbles with Anthropic's "When AI builds itself" blog post, but I appreciate them coming right out and saying this.

[dave kasten @David_Kasten]

Yesterday, I gave a talk where I said, "Hey, I'm more optimistic about AI policy than I was a year ago."

Things like seeing Senator Banks leading on RSI risks are a big part of why I feel that way.

[M A X @dafunnysun]

SCP 173 Reimagining

[Mitt Romney @MittRomney]

Our highest and most urgent national priority should be AI safeguards. The risks of AI weapons, pathogens, mass unemployment, surveillance, and even extinction must not continue to be largely ignored.

[R. Donald J. Gauvreau @cmSalticidae]

@JanayKinney @MaceAhWindu There are (at least) two angles. YMMV whether they matter; I think #1 is what usually concerns people.

1. Humans are holistic. Someone who'll break a promise to their partner might also break a promise to constituents.

2. Anything embarrassing is potential blackmail material.

[Connor Leahy @NPCollapse]

Right now, 20 British MPs are deciding which bills they'll introduce to Parliament.

Sir Stephen Fry just asked them to bring forward our bill to ban superintelligence!

Grateful to receive this strong endorsement from such a respected figure.

"Let the right thing be done."

[Harlan Stewart @HumanHarlan]

@DavidSacks Six months ago (the last time a journalist actually pressed him on it afaik), Altman said he still thinks superintelligent AI is the "greatest threat to the continued existence of humanity." But great to hear he thinks the jobs will be ok

https://t.co/mqBTGgBP9U

[Harlan Stewart @HumanHarlan]

We're gonna need a bigger graph

[Nate Soares ⏹️ @So8res]

"existential": academic, abstract, pearl-clutching
"risk": rare, unlikely, ignoring it is based
"x-risk": nerdy, jargon, lame

"extinction threat": oh you mean we might all actually fucking die

[AI StopWatch @AIStopWatch]

What will be the impact of AI industry super PACs?

"The takeaway here is that this year’s U.S. midterm elections are being aggressively shaped by different factions of the AI industry sometimes supporting the same candidates, sometimes different candidates, buying ads that don’t have anything to do with AI."

[R. Donald J. Gauvreau @cmSalticidae]

@goldfish_pond @lazy_grad @tenobrus realizing today that Sam Kriss is an LLM wrought in flesh and blood.

[Elizabeth Barnes @BethMayBarnes]

Sometimes people outside the field say things like “The AI situation can’t be that bad, there must be experts who are on top of it”. As “an expert”, I would like to be clear that we are *not* on top of it. Some key aspects of the situation IMO: