Shakeel

@lawrencenewport why don't you just join the Soho Society — then you can vote in their AGMs. £25 and everyone is eligible, if enough people we could completely reshape how it operates. x.com/ShakeelHashim/…

Done! Join the LFG: SOHO SOCIETY Let's fight the war on fun! lfgsohosociety.uk

yeah from a quick look nothing really great exists (maybe we'll write something!), but I think these links a) make the case for evals as a prerequisite for regulation and b) provide examples of AISI helping here: hamel.dev/notes/llm/eval… apolloresearch.ai/blog/apollo-is… arxiv.org/pdf/2403.13793 metr.org/blog/2023-09-2… arxiv.org/pdf/2305.15324

@ednewtonrex i will have a think and try to find something — from what I know there isn't a great summary anywhere though

I feel very uncomfortable about the AI Security Institute, and this article by its creator, ex-PM Rishi Sunak, encapsulates why: He frames it as a way of avoiding regulating AI. He says this categorically. “If politicians are blithe about the risks [of AI], they will vote for those who favour regulation.” And he makes it clear he thinks regulation would be bad: “western governments shouldn’t restrict innovation in the race against China.” This is a man who now works for both Anthropic and Microsoft (not disclosed in the article). Evaluating the safety of AI models is good. But not in place of regulating the technology itself, and the companies behind it. Besides, as he points out, companies only give the AI Security Institute access to their models voluntarily. This means the UK government relies on good relations with big tech, which in turn makes it even less likely to regulate. It is far from clear how much the AI Security Institute has actually achieved with its astronomical public funding. What *is* clear, though, is that the people who set it up see it as a way of ensuring *less* regulation of AI companies, not more. This is very bad news.

Actually on re-reading the Sunak article, I don’t think he’s making the claim you are. He’s saying that unless you have *some* regulation, there’ll inevitably be a backlash that will lead to overregulation, which will be counter productive. I think he’s right there (nuclear energy being a good example). And he says that if labs stop voluntarily complying we should make it mandatory.

@ednewtonrex I think the redteaming/jailbreaking work they do is better than what Pliny etc do, and importantly they can also redteam in natsec areas that literally only govt can red team. I think that’s a fair point re the Sunak article, I just think he’s wrong!

The evals stuff is the entire reason it exists, certainly not a bonus. I think you seriously underestimate how important evals are to figuring out catastrophic risk stuff (or how expensive it is to do stuff like this — look at US AISI, which is non-functional on its budget). Evals also massively strengthen the case for regulation: the post-Mythos desire to regulate is driven by evals making its capabilities more salient. (This is the whole METR theory of change, but they lack credibility compared to AISI). I also think that if and when we do regulate — which I think is absolutely going to happen in the next two years — we'll be very grateful to have spent a few years building up the infrastructure and skills to be able to do it. And regulation is much more politically feasible when the science of evals is well-established (the main talking point tech trade groups used against SB 315, for instance, was "no one knows what auditing looks like", being able to point to UK AISI's work to disprove that is really crucial). I agree we need mandatory regulation, but I think on net AISI's work and existence strengthens the case for regulation rather than undermines it.

I agree that AISI could be a good thing in principle, but I disagree that the bits of work you mention are clearly worth the funding. Finding jailbreaks is all well and good, but not vastly different from what the likes of Pliny do (for free). Besides, if this is deemed important, it feels like something that should be concretely regulated (requiring model access, audits etc.). I think there’s a good argument a voluntary setup gives people a sense of security without it really being there. I don’t think the evals work is worth the funding. Again, all well and good, but surely a bonus as opposed to the main thing you’d expect for these millions. IMO AISI’s main benefit is as a body that can share intel on model capabilities etc with the UK government (not just pitches that come from AI companies). In that regard I think it already performs a useful function. That alone wouldn’t justify the level of funding, though. Again, I don’t think its existence is clearly a bad thing. But I think there are potential flaws at the moment: (i) voluntary nature of engagements, which leads to (ii) chumminess with big tech, making it harder to regulate, (iii) Sunak’s admission in today’s article that AISI’s very existence diminishes the need to regulate, and (iv) a lack of transparency meaning it’s not clear what it’s achieving.

@AaronBastani anyone can become a member — let's take back control x.com/ShakeelHashim/…

The AGM of the Soho society. I’ve great respect for older people, they’re invariably wiser than those younger than them! But the median age in this country is 41, not 71. You can’t just create a sh*t country for young people and expect them to stay!

Anyone can become a member of Soho Society, by the way. Costs £25 and gets you the right to vote in their AGMs. I've just signed up.

@JacquesThibs why does it default to this, fascinating

@ShakeelHashim Unfortunately,

I'll take it

The Peter Thiel-funded startup Objection[.]ai — which was unveiled last month as a tool for rich people to pay money to have former spooks and AI models challenge journalists — appears to be defunct already.

I think @DavidSacks has shot himself in the foot. By consistently blocking AI regulation at the federal level, he's created a vacuum — one which states are now filling with increasingly stronger regulation of their own.

This applies to many many AI safety people who ought to know better

@ApriiSR I don’t really know how else to interpret this “seemingly human” line

@ShakeelHashim the whole scheme is unquestionably dumb and vile but i don't really think it's trying to push the idea that undocumented immigrants are not human in particular admittedly some people might interpret it as supporting that idea but i don't especially think that was the intent

The Trump admin using a fancily designed website to push the idea that undocumented immigrants are not human. Is this what Airbnb founder @jgebbia is spending his time on as Trump’s “Chief Design Officer”?

Illinois' new AI safety bill shows that the campaign to stop federal AI laws is backfiring. The longer Washington fails to regulate AI, the more room there is for states to pass stronger legislation — and raise the bar that an eventual federal framework will have to clear. The question now is whether accelerationists will finally realize this. transformernews.ai/p/the-campaign…

Really annoying not being able to control Cowork sessions from my phone, @AnthropicAI plz fix