This is a special post for quick takes by Alistair Stewart. Only they can create top-level comments. Comments here also appear on the Quick Takes page and All Posts page.
Sorted by Click to highlight new quick takes since:

Why aren't we engaging in direct action (including civil disobedience) to pause AI development?

Here's the problem:

Yudkowksy: "Many researchers steeped in these issues, including myself, expect that the most likely result of building a superhumanly smart AI, under anything remotely like the current circumstances, is that literally everyone on Earth will die."

Here's one solution:

FLI Open Letter: "all AI labs...immediately pause for at least 6 months the training of AI systems more powerful than GPT-4. This pause should be public and verifiable, and include all key actors. If such a pause cannot be enacted quickly, governments should step in and institute a moratorium."

Here's what direct action in the pursuit of that solution could look like (most examples are from the UK climate movement):

Picketing AI offices (this already seems to be happening!)

Mass non-disruptive protest

Strikes/walk-outs (by AI developers/researchers/academics)

Slow marches

Roadblocks

Occupation of AI offices

Performative vandalism of AI offices

Performative vandalism of art

Sabotage of AI computing infrastructure (on the model of ecotage)

Theory of change:

In the words of Martin Luther King Jr., activists seek to "create such a crisis and foster such a tension that a community...is forced to confront the issue". Activists create disruption, gain publicity, generate (moral) outrage, and set an agenda; they force people – civil society, companies, governments – to think about an idea they weren't previously thinking about. This in turn can shift the Overton window, enact social change, and lead to political/legislative/policy change – e.g. a government-enforced moratorium on AI development.

Final thoughts:

AI-focused direct action on the model of climate activism currently seems extremely neglected and potentially highly effective. As a problem, the threat from AI is plausibly both more important and more tractable than climate change: a government-enforced global moratorium on AI development seems easier to achieve than a government-enforced global moratorium on e.g. issuing new fossil fuel licences, because there is arguably less to lose and the companies who are developing AI are smaller in number and (currently?) less powerful than e.g. fossil fuel companies.

Some potential reasons why not were posted here:

https://www.lesswrong.com/posts/o8nPDZmmiLQhi6fwt/support-me-in-a-week-long-picketing-campaign-near-openai-s

I think this is potentially a unilateralist-curse-type situation, so people should be careful before they engage in these types of disruption.

Thanks for the link – very helpful. I'm surprised by how unpopular the suggestion of an OpenAI picket is on LW.

To be clear, is your suggestion that engaging in AI-focused direct action could lead to a unilateralist's curse-type situation in which one government (presumably a goodish actor) pauses AI development, leaving others (presumably worse actors) to develop AI more easily?

If we could create a global AI-focused movement that would pressure governments simultaneously into coordinating a multilateral moratorium on development, would you support that?

No, I mean unilateralist curse in a different way:

Let's say lots of LessWrong people think about whether to go blockade the OpenAI offices while holding signs and yelling at people that AI will kill everyone. Most of the people conclude that this is a bad idea because there could be large negative consequences (potentially: OpenAI engineers becoming hostile to safety efforts, polarization, spreading an inaccurate picture of the problem, the wrong memes (terminator, killer robots) go mainstream, makes AI Alignment look crazy, etc.). However, one small minority of LessWrong people could think otherwise and that the effects are positive, and go on and do blockade the OpenAI offices.

What one should do in such a situation: coordinate with a larger group of people, think more carefully whether this is a good idea and listen to other people's argument, etc.

I see – and I presume you would agree with the majority of OpenAI people in this situation (i.e. direct action is a bad idea)?

Would you say the same thing about direct action taken against fossil fuel companies?

I think protesting/blocking fossil fuel companies is different and less of a unilateralist curse situation. For example, there is wide elite/expert agreement that more CO2 in the atmosphere is bad. We do not have that for the extinction of humanity due to AI. There also have been many protests against fossil fuel already, so additional protest is less likely to cause serious downsides or set the tone for future attempts to solve the problem. The nature of the problem is also different: incompetent political solutions to solve global warming often still help reduce CO2 somewhat, but the same might not be true for AI Notkilleveryoneism.

I am not sure whether "direct action" (imo a terrible name btw if the theory of change is indirect) against AI would be a good idea but lean against it currently.

For example, there is wide elite/expert agreement that more CO2 in the atmosphere is bad. We do not have that for the extinction of humanity due to AI.

We don't need to believe that AI will lead to human extinction to advocate for a moratorium on AI development. Karnofsky outlines a number of ways in which TAI could lead to global catastrophe here; and this 2021 survey of 44 AI risk researchers found the median estimate of existential risk was 32.5%. The risk from AI is a huge problem.

There also have been many protests against fossil fuel already, so additional protest is less likely to cause serious downsides or set the tone for future attempts to solve the problem.

Do you think that climate protest is more harmful than helpful when it comes to solving the climate crisis?

The nature of the problem is also different: incompetent political solutions to solve global warming often still help reduce CO2 somewhat, but the same might not be true for AI Notkilleveryoneism.

This is a good point – but that's an argument for competent political solutions, not no political solutions (which is roughly what we have at the moment I think?).

I am not sure whether "direct action" (imo a terrible name btw if the theory of change is indirect) against AI would be a good idea but lean against it currently.

Learn more here!

Curated and popular this week
Ben_West🔸
 ·  · 1m read
 · 
> Summary: We propose measuring AI performance in terms of the length of tasks AI agents can complete. We show that this metric has been consistently exponentially increasing over the past 6 years, with a doubling time of around 7 months. Extrapolating this trend predicts that, in under a decade, we will see AI agents that can independently complete a large fraction of software tasks that currently take humans days or weeks. > > The length of tasks (measured by how long they take human professionals) that generalist frontier model agents can complete autonomously with 50% reliability has been doubling approximately every 7 months for the last 6 years. The shaded region represents 95% CI calculated by hierarchical bootstrap over task families, tasks, and task attempts. > > Full paper | Github repo Blogpost; tweet thread. 
 ·  · 2m read
 · 
For immediate release: April 1, 2025 OXFORD, UK — The Centre for Effective Altruism (CEA) announced today that it will no longer identify as an "Effective Altruism" organization.  "After careful consideration, we've determined that the most effective way to have a positive impact is to deny any association with Effective Altruism," said a CEA spokesperson. "Our mission remains unchanged: to use reason and evidence to do the most good. Which coincidentally was the definition of EA." The announcement mirrors a pattern of other organizations that have grown with EA support and frameworks and eventually distanced themselves from EA. CEA's statement clarified that it will continue to use the same methodologies, maintain the same team, and pursue identical goals. "We've found that not being associated with the movement we have spent years building gives us more flexibility to do exactly what we were already doing, just with better PR," the spokesperson explained. "It's like keeping all the benefits of a community while refusing to contribute to its future development or taking responsibility for its challenges. Win-win!" In a related announcement, CEA revealed plans to rename its annual EA Global conference to "Coincidental Gathering of Like-Minded Individuals Who Mysteriously All Know Each Other But Definitely Aren't Part of Any Specific Movement Conference 2025." When asked about concerns that this trend might be pulling up the ladder for future projects that also might benefit from the infrastructure of the effective altruist community, the spokesperson adjusted their "I Heart Consequentialism" tie and replied, "Future projects? I'm sorry, but focusing on long-term movement building would be very EA of us, and as we've clearly established, we're not that anymore." Industry analysts predict that by 2026, the only entities still identifying as "EA" will be three post-rationalist bloggers, a Discord server full of undergraduate philosophy majors, and one person at
Thomas Kwa
 ·  · 2m read
 · 
Epistemic status: highly certain, or something The Spending What We Must 💸11% pledge  In short: Members pledge to spend at least 11% of their income on effectively increasing their own productivity. This pledge is likely higher-impact for most people than the Giving What We Can 🔸10% Pledge, and we also think the name accurately reflects the non-supererogatory moral beliefs of many in the EA community. Example Charlie is a software engineer for the Centre for Effective Future Research. Since Charlie has taken the SWWM 💸11% pledge, rather than splurge on a vacation, they decide to buy an expensive noise-canceling headset before their next EAG, allowing them to get slightly more sleep and have 104 one-on-one meetings instead of just 101. In one of the extra three meetings, they chat with Diana, who is starting an AI-for-worrying-about-AI company, and decide to become a cofounder. The company becomes wildly successful, and Charlie's equity share allows them to further increase their productivity to the point of diminishing marginal returns, then donate $50 billion to SWWM. The 💸💸💸 Badge If you've taken the SWWM 💸11% Pledge, we'd appreciate if you could add three 💸💸💸 "stacks of money with wings" emoji to your social media profiles. We chose three emoji because we think the 💸11% Pledge will be about 3x more effective than the 🔸10% pledge (see FAQ), and EAs should be scope sensitive.  FAQ Is the pledge legally binding? We highly recommend signing the legal contract, as it will allow you to sue yourself in case of delinquency. What do you mean by effectively increasing productivity? Some interventions are especially good at transforming self-donations into productivity, and have a strong evidence base. In particular:  * Offloading non-work duties like dates and calling your mother to personal assistants * Running many emulated copies of oneself (likely available soon) * Amphetamines I'm an AI system. Can I take the 💸11% pledge? We encourage A