Perhaps consider support for my "Beer And Lottery Tickets" ngo where I, personally, will invest in buying myself beer and lottery tickets. A year long pilot program will prove the efficacy.
Zvi or anyone else reading. Do you know or have contacts with search engineers at Google who’ve caused major server side issues. Are you either able to talk to them and ask them to talk to me or raise the issues on your blog? This would be helpful
This post is great, but I think the way you use the term "Funding Needed" is misleading
> They roughly mean this, to the best of my knowledge:
> Low: Could likely be fully funded with less than ~$250k.
> Medium: Could plausibly be fully funded with between ~$250k and ~$2 million.
> High: Could probably make good use of more than ~$2 million.
This means that a small org, which is in desperate need for money (e.g. would have to shut down soon if they don't get donations), but that has a small total budget (anything below $250k) will be ranked as "Funding Needed: Low".
This means that donation oportunity that could really make a diffrence, may be overlooked because people skimming this post doesn't think the money is needed.
> I loved the simple core concept of a ‘catered hotel’ where select people can go to be supported in whatever efforts seem worthwhile. They are now broadening their approach, scaling up and focusing on logistical and community supports, incubation and a general infrastructure play on top of their hotel. This feels less unique to me now and more of a typical (EA UK) community play, so you should evaluate it on that basis.
Having got a read on the ground, the previous value proposition is still very much going strong, and there are no plans to remove that. As I understand it, the basic mid term plan is to have a mix of residencies and live-in mentor-ish people who are doing their own things, but add in some drives to pitch seasons for people with overlapping interests so that there's greater opportunity for cross-pollination. There a bunch of other things that could come together as extras, but the team here is keen to keep the things the community knows and loves.
More significantly, the read on the ground I get is extremely positive, much more than previous years now that it's got active full-time management by someone with relevant experience and drive. Multiple people have said things at least as positive as "this place is life changingly amazing, unblocked me from years long dips, helped me get way more productive, etc", and it's pretty clear that the arc of things is spinning up towards people who have much more and better outputs than in previous years.
Attila's drive to clear the backlog of work needed to get the EA hotel organised, upgraded as a living space, and increased intentionality around selection+getting higher deal-flow so everyone here is agentic and competent is causing this place to spin up an increasing amount of momentum. I honestly think the EA hotel is one of the best EV places to support in the AI safety space right now. My guess is within 3 months we will see several counterfactual outputs which would individually justify the relatively small budget of ~$350k/year to support 20-30 people with a low-hassle and awesome environment.
(CoI: visiting and have friends here, but confident that I would make the same claims if this was not true)
> Focus: Making YouTube videos about AI safety, starring Rob Miles
> Leader: Rob Miles
> Funding Needed: Low
> Confidence Level: High
> I think these are pretty great videos in general, and given what it costs to produce them we should absolutely be buying their production. If there is a catch, it is that I am very much not the target audience, so you should not rely too much on my judgment of what is and isn’t effective video communication on this front, and you should confirm you like the cost per view.
These are two separate-ish projects, Rob Miles makes videos, and Rob Miles is the project owner of AISafety.info mostly in an advisory role. Rob Miles personally is not urgently in need of funding afaik, but will need to reapply soon. AIsafety.info is in need of funding, and recently had a funding crunch which caused several staff members to have to drop off payroll. AISafety.info writers have helped Rob with scriptwriting some, but it's not their main focus. Donate link for AI Safety Info.
I can't speak precisely for everyone in Pause AI, but I do fear there are some misunderstandings here (predicted and downstream of a deliberately simple branding choice) where the typical Pause AI mindset is much less distant from yours than you suggest.
We are addressing this more fully with professional help, but here is my personal and individually expressed heads up for you and yours.
Three things:
* One movement, many chapters
* Funding opportunity is larger
* Pause is a brand
> Pause AI and Pause AI Global
These are the same thing. I figured maybe you meant to link Pause AI US first?
Well, either way: the US is the largest and most critical national chapter. They specialize very well to the political (state, congress, senate) opportunities in the US and to the lab presence ones. Holly and Felix et al rock.
But national (and city) chapters are a scaling mechanism. This is standard activist practice. The UK, German, French, Italian Canadian and Australian chapters are also large and at time of writing we list eighty-five different city chapters on https://pauseai.info/communities
There is a single Pause AI movement. Act locally, coordinate globally.
> Focus: Advocating for a pause on AI, including via in-person protests
Yes. A resource for and composed of the general public. We want to inform everyone about extinction risk level AI and help them mobilize and advocate against a race that can kill us.
> Funding Level: Low
Everyone will always take more, but believe this was a category error, we already today spend at a level you'd class Medium.
> I don’t think pausing right now is a good idea.
That's... such a weird hypothetical.
One would need a magic wand. And I'd really want to scrutinize the international agreement and enforcement mechanisms that manifested when it was waved, and debate the settings.
> I think we should be working on the transparency, state capacity, technical ability and diplomatic groundwork to enable a pause in case we need one
Same. The pause dial needs built, and then dialed back some. Starting earlier on both building and using the dial is better.
> but that it is too early to actually try to implement one.
I don't get what this distinction means in practice.
We need to *work* on it now. When we do get any dial, most members would turn it quite hard counter-clockwise straight away, but I think you would too?
Would I dial improvements in capabilities that enable super-intelligence as built today to *zero* today if I could, whatever that could mean?
I guess... probably? It seems like appropriate caution. But I have epistemic humility, I expect to learn more, and I grok reality enough to know what's practical shorter term.
> I do think that if you believe we should pause? Then you should say that we should pause. [...] If you disagree, you have better options.
The tent is bigger than that, though.
If you believe we should slow down a lot, as soon as we can, and would want to pause before "anyone builds an artificial superintelligence using anything remotely like current techniques, based on anything remotely like the present understanding of AI" because you believe there's a *decent chance* everyone then dies, you are not at all out of place in Pause AI.
The brand is simpler and dumber and more iconic than that because you don't even get five words, let alone seventy.
And sure, do fund us if that's how you roll. But donating time and energy matters more.
For ALTER, I think the description misses most of what we're doing. The vast majority of our attention is on international policy for AI, and to some extent we're attempting to leverage the fact that we are in Israel to accomplish some of those things.
Our most recent update, from mid-year, has a bunch more detail, though we'll have another update in the next month or so:
Are you thinking specifically of hard dollars to political campaigns? I’d love to better understand how those opportunities compare to what Zvi’s included here.
This articole comes at the perfect time! Such a brilliant resorce for finding worthy causes. It's inspiring to see so many quality initiatives, but also a bit overwhelming how much good work is needed.
Another thought: Every time I read ‘Funding Needed: Low’, I instinctively think ‘this project doesn't need money’ and have to remind myself of the explanation somewhere at the top of the post. I would suggest just putting the $ range there.
For people who are tax residents in Germany or Switzerland: Effektiv Spenden has regranted to some of the listed organisations in the past (e.g., The Future Society, SaferAI, or METR) and would be open to do this again provided the total amount per org is above €20k. Donations via us would be tax-deductible (you can claim back around 40% of your donation from your taxes). Feel free to reach out at info@effektiv-spenden.org.
Also, plug: For smaller amounts, we offer tax-deductible donations via our Giving Fund: Ensuring Safe AI, which is advised by Longview Philanthropy: https://effektiv-spenden.org/en/ki-kontrollieren/.
Perhaps consider support for my "Beer And Lottery Tickets" ngo where I, personally, will invest in buying myself beer and lottery tickets. A year long pilot program will prove the efficacy.
Zvi or anyone else reading. Do you know or have contacts with search engineers at Google who’ve caused major server side issues. Are you either able to talk to them and ask them to talk to me or raise the issues on your blog? This would be helpful
This post is great, but I think the way you use the term "Funding Needed" is misleading
> They roughly mean this, to the best of my knowledge:
> Low: Could likely be fully funded with less than ~$250k.
> Medium: Could plausibly be fully funded with between ~$250k and ~$2 million.
> High: Could probably make good use of more than ~$2 million.
This means that a small org, which is in desperate need for money (e.g. would have to shut down soon if they don't get donations), but that has a small total budget (anything below $250k) will be ranked as "Funding Needed: Low".
This means that donation oportunity that could really make a diffrence, may be overlooked because people skimming this post doesn't think the money is needed.
Podcast episode for this post: 2 hours of nonprofity goodness.
https://open.substack.com/pub/dwatvpodcast/p/the-big-nonprofits-post-2025
FYI The URL for AI Safety Cape Town doesn't work anymore. The project has expanded itself to AI Safety South Africa: https://www.aisafetysa.com/
> CEEALAR / EA Hotel
> I loved the simple core concept of a ‘catered hotel’ where select people can go to be supported in whatever efforts seem worthwhile. They are now broadening their approach, scaling up and focusing on logistical and community supports, incubation and a general infrastructure play on top of their hotel. This feels less unique to me now and more of a typical (EA UK) community play, so you should evaluate it on that basis.
Having got a read on the ground, the previous value proposition is still very much going strong, and there are no plans to remove that. As I understand it, the basic mid term plan is to have a mix of residencies and live-in mentor-ish people who are doing their own things, but add in some drives to pitch seasons for people with overlapping interests so that there's greater opportunity for cross-pollination. There a bunch of other things that could come together as extras, but the team here is keen to keep the things the community knows and loves.
More significantly, the read on the ground I get is extremely positive, much more than previous years now that it's got active full-time management by someone with relevant experience and drive. Multiple people have said things at least as positive as "this place is life changingly amazing, unblocked me from years long dips, helped me get way more productive, etc", and it's pretty clear that the arc of things is spinning up towards people who have much more and better outputs than in previous years.
Attila's drive to clear the backlog of work needed to get the EA hotel organised, upgraded as a living space, and increased intentionality around selection+getting higher deal-flow so everyone here is agentic and competent is causing this place to spin up an increasing amount of momentum. I honestly think the EA hotel is one of the best EV places to support in the AI safety space right now. My guess is within 3 months we will see several counterfactual outputs which would individually justify the relatively small budget of ~$350k/year to support 20-30 people with a low-hassle and awesome environment.
(CoI: visiting and have friends here, but confident that I would make the same claims if this was not true)
> AI Safety Info (Robert Miles)
> Focus: Making YouTube videos about AI safety, starring Rob Miles
> Leader: Rob Miles
> Funding Needed: Low
> Confidence Level: High
> I think these are pretty great videos in general, and given what it costs to produce them we should absolutely be buying their production. If there is a catch, it is that I am very much not the target audience, so you should not rely too much on my judgment of what is and isn’t effective video communication on this front, and you should confirm you like the cost per view.
These are two separate-ish projects, Rob Miles makes videos, and Rob Miles is the project owner of AISafety.info mostly in an advisory role. Rob Miles personally is not urgently in need of funding afaik, but will need to reapply soon. AIsafety.info is in need of funding, and recently had a funding crunch which caused several staff members to have to drop off payroll. AISafety.info writers have helped Rob with scriptwriting some, but it's not their main focus. Donate link for AI Safety Info.
Donate link here: https://www.every.org/aisafetyinfo?donateTo=aisafetyinfo#/donate/card
Zvi, thanks for the vote of confidence!
I can't speak precisely for everyone in Pause AI, but I do fear there are some misunderstandings here (predicted and downstream of a deliberately simple branding choice) where the typical Pause AI mindset is much less distant from yours than you suggest.
We are addressing this more fully with professional help, but here is my personal and individually expressed heads up for you and yours.
Three things:
* One movement, many chapters
* Funding opportunity is larger
* Pause is a brand
> Pause AI and Pause AI Global
These are the same thing. I figured maybe you meant to link Pause AI US first?
Well, either way: the US is the largest and most critical national chapter. They specialize very well to the political (state, congress, senate) opportunities in the US and to the lab presence ones. Holly and Felix et al rock.
But national (and city) chapters are a scaling mechanism. This is standard activist practice. The UK, German, French, Italian Canadian and Australian chapters are also large and at time of writing we list eighty-five different city chapters on https://pauseai.info/communities
There is a single Pause AI movement. Act locally, coordinate globally.
> Focus: Advocating for a pause on AI, including via in-person protests
Yes. A resource for and composed of the general public. We want to inform everyone about extinction risk level AI and help them mobilize and advocate against a race that can kill us.
> Funding Level: Low
Everyone will always take more, but believe this was a category error, we already today spend at a level you'd class Medium.
> I don’t think pausing right now is a good idea.
That's... such a weird hypothetical.
One would need a magic wand. And I'd really want to scrutinize the international agreement and enforcement mechanisms that manifested when it was waved, and debate the settings.
> I think we should be working on the transparency, state capacity, technical ability and diplomatic groundwork to enable a pause in case we need one
Same. The pause dial needs built, and then dialed back some. Starting earlier on both building and using the dial is better.
> but that it is too early to actually try to implement one.
I don't get what this distinction means in practice.
We need to *work* on it now. When we do get any dial, most members would turn it quite hard counter-clockwise straight away, but I think you would too?
Would I dial improvements in capabilities that enable super-intelligence as built today to *zero* today if I could, whatever that could mean?
I guess... probably? It seems like appropriate caution. But I have epistemic humility, I expect to learn more, and I grok reality enough to know what's practical shorter term.
> I do think that if you believe we should pause? Then you should say that we should pause. [...] If you disagree, you have better options.
The tent is bigger than that, though.
If you believe we should slow down a lot, as soon as we can, and would want to pause before "anyone builds an artificial superintelligence using anything remotely like current techniques, based on anything remotely like the present understanding of AI" because you believe there's a *decent chance* everyone then dies, you are not at all out of place in Pause AI.
The brand is simpler and dumber and more iconic than that because you don't even get five words, let alone seventy.
And sure, do fund us if that's how you roll. But donating time and energy matters more.
https://pauseai.info/join
(Other members, do say where I'm wrong!)
For ALTER, I think the description misses most of what we're doing. The vast majority of our attention is on international policy for AI, and to some extent we're attempting to leverage the fact that we are in Israel to accomplish some of those things.
Our most recent update, from mid-year, has a bunch more detail, though we'll have another update in the next month or so:
https://forum.effectivealtruism.org/posts/5HPzmrHzQCsqo6ZDE/alter-israel-mid-2025-semiannual-update
For what it's worth, I'd prefer retail donors donate to promising AI politics opportunities!
Are you thinking specifically of hard dollars to political campaigns? I’d love to better understand how those opportunities compare to what Zvi’s included here.
This articole comes at the perfect time! Such a brilliant resorce for finding worthy causes. It's inspiring to see so many quality initiatives, but also a bit overwhelming how much good work is needed.
Another thought: Every time I read ‘Funding Needed: Low’, I instinctively think ‘this project doesn't need money’ and have to remind myself of the explanation somewhere at the top of the post. I would suggest just putting the $ range there.
For people who are tax residents in Germany or Switzerland: Effektiv Spenden has regranted to some of the listed organisations in the past (e.g., The Future Society, SaferAI, or METR) and would be open to do this again provided the total amount per org is above €20k. Donations via us would be tax-deductible (you can claim back around 40% of your donation from your taxes). Feel free to reach out at info@effektiv-spenden.org.
Also, plug: For smaller amounts, we offer tax-deductible donations via our Giving Fund: Ensuring Safe AI, which is advised by Longview Philanthropy: https://effektiv-spenden.org/en/ki-kontrollieren/.