Written in my personal capacity.

Quick summary: over the past couple of months, I've been spending my free time working with some collaborators to figure out the best ways to donate money and share our takes with prospective donors. We differ from existing orgs in that we're willing to evaluate and recommend time-sensitive opportunities, and we don't shy away from political donation opportunities. If your donation budget is over $30,000/year and you're interested in our takes, please email me!

Introduction

Last month, I wrote a blog post where I analyzed the cost-effectiveness of donating to Alex Bores, an AI safety champion running for Congress in New York, and opined that for most donors, donating to Alex Bores is the most effective use of money for making the long-term future go well.

(I then hastily wrote a similar blog post about AI safety champion Scott Wiener, an AI safety champion who coincidentally announced his run for Congress just two days later.)

I was encouraged by the positive response to my posts: it turned out that many people found them helpful! But that also raised the question: why isn't anyone else doing this? In a community of people who care a ton about the most effective ways to donate money, why wasn't anyone else set up to make similarly detailed cost-effectiveness analyses? Isn't that, like, the specialty of multiple organizations in this space?

Well... no, actually! Existing organizations couldn't have done this, for at least two reasons:

• They generally move slowly, carefully vetting donations before recommending grants.^[1]
• They mostly steer clear of politics, and steer fully clear of recommending donations to specific political candidates, in part for legal reasons.

And so, I think there's a gap in the landscape of assessing donation opportunities. Over the past couple of months, I have been using my spare time to try to fill this gap. I have been working with Zach Stein-Perlman and a few other collaborators on this project.

What we've been up to

Our biggest project over the past month has been putting together a guide on the best uses of money for making the long-term future go well. This involves researching the donation opportunities that we think might be among the best, and doing a cost-effectiveness analysis of each one, carefully but expediently.

Overall, we currently think that most of the best opportunities for making the long-term future go well lie in a few buckets, including:

• Supporting the political campaigns of potential AI safety champions, such as Alex Bores or Scott Wiener.
• Steering toward worlds where the people in charge of making the most important decisions around the time of AGI are competent and wise.
• Educating key decision-makers on catastrophic risks from advanced AI.

However, we are also evaluating many opportunities that don't fall in these buckets, and will continue to do so.

Should you take our advice?

One good heuristic for whether you should take our advice is whether you like the sort of analysis I did in my blog post about Alex Bores. If you're like, "Yes, I'd love to have more of this sort of analysis, or defer to the sorts of people who do this kind of analysis," I think you'd benefit from our recommendations.

If so, and if you're a medium or large donor^[3] (e.g. your donation budget at least $30,000/year), please reach out! You can find my email address here.

1. ^{^}

   This is admirable, of course, but there are important trade-offs, and being slow isn't always the right move.

2. ^{^}

   See e.g. here for some discussion of career costs of political donations.

3. ^{^}

   We might have some outputs for smaller donors as well, though this isn't a high priority for us at the moment. But our current #1 recommendation is donating $7,000 to Alex Bores, and our #2 recommendation is donating $7,000 to Scott Wiener, which already gets you to $14,000. (Of course, these recommendations are subject to change.)