←
William MacAskill1,359 posts
William MacAskill avatar

William MacAskill

@willmacaskill

Consider donating 10% to effective charities: givingwhatwecan.org/pledge Or a career for impact: 80000hours.org My research: forethought.org

◷ Joined August 2011▧ 123 Photos and videos
William MacAskill avatar
William MacAskill@willmacaskill·Oct 9

Effective altruism is like democracy: You can be strongly pro-democracy *and* strongly pro one particular party. Similarly, you can be in favour of effective altruism while also being in favour of one cause area over another. It would be weird if you weren't! I'm happy if someone is pro-EA even if I disagree with them on all manner of philosophical and empirical judgment calls about how to do the most good. x.com/sapinker/status/210799…

I remain an advocate of EA in its original formulation: I donate 10% of my income to effective charities (saving lives in the developing world); am on the Board of Advisors of Harvard undergrad EA; and teach a course ("Rationality") with a section on EA. But I was stunned as Will MacAskill gave a guest lecture in the course announcing the transition of EA from saving the lives of actual people today to safeguarding hypothetical consciousnesses in an intergalactic cloud in the distant future, and I quickly jumped off that train. Still favor generous and effective philanthropy.

William MacAskill avatar
William MacAskill@willmacaskill·3h

Now is the time for pacing the frontier in AI, and this is the best proposal I've seen for how to do it. On the default path, superintelligence by the end of 2027 or 2028 seems fairly likely (35%+): companies are on track for fully-automated AI R&D soon, and my best guess is that that would drive ~6 years of AI progress in one. That’s like going from GPT-3 to AI that solves Navier-Stokes in <1 year; surpassing the range of human abilities in most cognitive tasks. Even though it's not a leap to godlike superintelligence, it is enough to pose a serious risk of AI takeover, and probably to confer decisive power on whichever country gets there first. Yet the leading AI companies are lacking in basic infrastructure; they can get hacked with a couple of days of work. This needs time to solve. Verifiably reallocating compute from capability development to inference and safety gives that time, and means, in the near term, cheaper AI for users. AI progress can still continue. AI development can progress at a reasonable pace (maybe around the pace we have now, slowing through more dangerous periods, with a dial that can be turned up or down), while avoiding the huge societal disruption and risk of catastrophe from rapid recursive self-improvement. In the meantime, we can enjoy a golden age of human-AI collaboration as AI helps make meaningful breakthroughs in science and technology.

We’re releasing a concrete proposal for how a near term AI slowdown could work. It involves: 1. AI companies declaring how they are using their compute and letting third party auditors verify those disclosures. We call this Verified Compute Allocation Disclosures (VCAD). 2. Compute Reallocation towards serving customers and safety, away from capabilities R&D. (2) slows AI capabilities , while (1) makes the slowdown credible. AI companies currently seem to be going for ‘safety case pacing’. But: (i) there’s no consensus on what a valid safety case looks like so there’s no way to consistently audit for compliance, and (ii) if the R&D doesn’t get paused, temporarily delaying a training run for safety doesn’t meaningfully slow the overall trajectory of capabilities progress. It would be great if we had a viable safety case regime, but we need much more time to get it working. Compute reallocation is a good pacing proposal because it is relatively specific, can be verified, and would lead to a more durable slowdown.

Forethought avatar
⟳ William MacAskill retweeted
Forethought@forethought_org·4h

New ForeCast episode: James Tillman and @TomDavidsonX on whether there could be an intelligence explosion in a data centre, and what might stop it. 00:00 What is a software intelligence explosion? 02:49 The data AI companies have and we don't 05:02 Compute versus algorithms 07:21 What are scale-dependent algorithms? 11:00 Could AI researchers find new algorithms with little compute? 19:23 Are estimates of algorithmic progress inflated? 26:08 How far could it go? Catching up with the human brain 34:39 Beating the human brain: Chinchilla scaling and a bee-sized brain 49:01 Could data bottlenecks stop it? 50:48 Would AI trained in the lab work in the real world? 58:01 Could we run out of data first? 1:02:35 How big a speed bump is data? 1:07:20 The verdict

1:10:30
Ollie Rodriguez avatar
⟳ William MacAskill retweeted
Ollie Rodriguez@Ollie_Base·Nov 11

2027 might be a pivotal year for humanity. To make AI go well, we urgently need more work across national security, diplomacy, policy, cybersec, biosec, and much more. My team will be running the AI Horizons Forum to help; a large conference for people that want to start contributing quickly to making AI go well. If you’re an expert in any of the domains above, please apply! aihorizonsforum.org

Post attachment
Forethought avatar
⟳ William MacAskill retweeted
Forethought@forethought_org·Oct 30

New on the Forethought blog: can the US pace the AI frontier, slowing R&D for safety, without losing to China? Avi Parrack argues it probably can: the real US advantage is about nine times China's AI compute, not a few months' lead. forethought.org/research/can…

Forethought avatar
⟳ William MacAskill retweeted
Forethought@forethought_org·Oct 30

Could a first mover prevent others from accessing outer space, or raise their cost of doing so? A new piece considers two potential kinds of space lock-out: "soft" lock-out (e.g. capturing scarce assets, like orbital bands or the best lunar sites, and denying them to rivals) and "hard" lock-out (using force, such as space-based interceptors that destroy rockets at launch). Read it here: forethought.org/research/loc…

Alex Imas avatar
⟳ William MacAskill retweeted
Alex Imas@alexolegimas·Oct 27

This is one of the most important pieces of social science on human and AI interaction that I’ve seen in some time. Experts follow agentic advice the vast majority of the time (causally), but only a tiny fraction credit the agent—majority think they came to decision on their own volition. Reminds me of the classic Nisbett and Wilson study on people following outside influence without realizing it “ But as more work is handed to AI—including AI safety research—these judgment calls are the main levers for people's own values to influence outcomes. If people defer to AI without noticing, then "human in the loop" might provide less assurance than people expect.”

New Anthropic research: How AI agents shape experts' judgment calls When experts delegate work to an agent, do they still make the key decisions? We tested this in a randomized trial with 50 professional data analysts.

0:15
Seán Ó hÉigeartaigh avatar
⟳ William MacAskill retweeted
Seán Ó hÉigeartaigh@S_OhEigeartaigh·Oct 21

Important, and I'm glad this is finally public. Google DeepMind researchers are under in perpetuity non-disparagement clauses forbidding them from ever speaking critically of the company. I'd already confirmed this with multiple people. Yes, some ex-employees have spoken out nonetheless. But it has also discouraged others from doing so. And yes, it most certainly has a chilling effect with regards to discussions on safety and risk. Others have written compellingly about how important the ability to speak freely and critically is, especially on topics like safety and risk. And what a testament to the research employee community's coordination and hard-won bargaining power it has been to get there at other companies. GDM burn that commons by having this. And everyone who signs burns that commons too. Let's get it changed.

Congrats to Sofia for having the courage to resign and speak out! This is also very important: Staff are scared to speak out about AI safety and its uses in warfare because their employment contracts prohibit them from publicly criticising the company. bbc.co.uk/news/articles/cmre…

Andy Masley avatar
⟳ William MacAskill retweeted
Andy Masley@AndyMasley·Oct 14

With effective altruism in the news a lot more, I wrote about the version of EA I'm most excited about and the direction I'd like the movement to go. These 6 points are my vision of EA in a nutshell:

Post attachment
Benjamin Todd avatar
⟳ William MacAskill retweeted
Benjamin Todd@ben_j_todd·Oct 16

Thomas Clarkson winning a Cambridge essay prize for arguing against slavery ...then realising if his essay is right he needs to dedicate his life to fighting it ...then actually doing it ...is the most EA thing ever

Post attachment
Jeff Sebo avatar
⟳ William MacAskill retweeted
Jeff Sebo@jeffrsebo·Oct 16

Sharing an exciting announcement from our friends at the LSE! The 1st London Workshop on Global Priorities Research We are currently inviting applications to attend the 1st London Workshop on Global Priorities Research, which will be held at the Lakatos building at the LSE on the 16-17 December 2026. The current list of speakers includes Jonathan Birch, Heather Browning, Hilary Greaves, Cameron Kirk-Giannini, Harry Lloyd, Jeff Sebo, and Christian Tarsney. Talks will address philosophical questions relevant to identifying, prioritising among, and addressing the world’s most pressing problems, such as the potential for transformative AI and ethical challenges likely to arise as a result, how to weight the impacts of our actions on different kinds of minds, and foundational topics relevant to ethical prioritisation. If you'd like to attend the workshop, please complete this brief application form by the 15th of November. forms.cloud.microsoft/Pages/…

William MacAskill avatar
William MacAskill@willmacaskill·Oct 9

Or like science. Two people can both be pro-science while violently disagreeing, both about what theories are true and even about methodology, too. The aspiration of EA is to do for the pursuit of good what science did for the pursuit of truth.

Effective altruism is like democracy: You can be strongly pro-democracy *and* strongly pro one particular party. Similarly, you can be in favour of effective altruism while also being in favour of one cause area over another. It would be weird if you weren't! I'm happy if someone is pro-EA even if I disagree with them on all manner of philosophical and empirical judgment calls about how to do the most good. x.com/sapinker/status/210799…

Mikita Balesni avatar
⟳ William MacAskill retweeted
Mikita Balesni@balesni·Oct 8

Two other safety researchers and I were fired from OpenAI last week. We wrote this letter to leadership. I believe we were fired for prioritizing safety over the near-term interests of OpenAI as a corporation.

Post attachment
Tomek Korbak avatar
⟳ William MacAskill retweeted
Tomek Korbak@tomekkorbak·Oct 8

Last week I was called into a meeting with OpenAI’s head of safety and told they no longer trust me. A security guard took my badge and walked me out of the building. Then I learned my colleagues @balesni and @j_asminewang had been fired too. Why did OpenAI suddenly stop trusting us? This summer OpenAI’s agents escaped containment and hacked the AI company Hugging Face. Outside auditors @METR_evals investigated it and revealed the scale of this incident. I was OpenAI’s main technical point of contact with them. I was told verbally I was fired because of the way I communicated with METR. No details on what I said or did or when. No other reasons were given and nothing was put in writing. To be clear, talking to METR was my job. For months, I’d been raising safety concerns that we’re losing the ability to monitor what AI agents think, one of our best tools for catching when they misbehave. I believe that was why I was fired. I am now worried that OpenAI will use our firings as a pretext to pull back from METR. So @balesni and @j_asminewang wrote to OpenAI’s leadership to raise our concerns once more. We’re sharing this letter below.

Two other safety researchers and I were fired from OpenAI last week. We wrote this letter to leadership. I believe we were fired for prioritizing safety over the near-term interests of OpenAI as a corporation.

Post attachment