by Brian Tomasik
First written: 9 Sept. 2013; last update: 1 Nov. 2014


For more current thoughts on where to donate, see "My Donation Recommendations".

As of 2014, I'm working for the Foundational Research Institute, exploring a variety of questions about how best to reduce suffering in the future. I'm also supporting Animal Ethics, a charity raising concern for animal suffering, including by animals in nature. I currently also endorse donations to the Machine Intelligence Research Institute (MIRI). I donated to several different organizations in the past, some of which I still support and some of which I now oppose.

Current projects

Right now I'm collaborating with friends in Switzerland on research about how best to reduce suffering from a broad perspective. I have high uncertainty not just about what the best project is for reducing suffering but even whether many of the conventional approaches are actually net positive or negative. There are multitude "foundational" issues that need to be explored to better shine light on these concerns. I think the best thing suffering reducers can do now is take a step back and study these questions for the next several years, before focusing too much on concrete activist projects. As Nick Bostrom says: "If we have overlooked even just one [crucial] consideration, then all our best efforts might be for naught---or less. When headed the wrong way, the last thing needed is progress."

That said, I'm also supporting and donating small amounts to a charity called Animal Ethics that seems like a relatively safe bet for being an important cause. Animal Ethics hopes to realize the dream of Yew-Kwang Ng in igniting the study of "welfare biology" by ecologists, ethologists, environmental scientists, and philosophers. We hope to encourage students and academics to take on important questions about how human actions affect wild-animal suffering. In addition, we hope to build interest in this topic among members of the animal-rights community and encourage them to examine their prior assumptions about life in the wild with unbiased eyes. Animal Ethics is a tax-deductible 501(c)(3) public charity in the US and has charitable status in Spain.

Other organizations I may endorse

As I discuss more in the next section, my views on MIRI have fluctuated back and forth, back and forth between MIRI being net positive and net negative for suffering reducers. I now think MIRI is likely net beneficial, because MIRI's work on friendliness has the potential to benefit all value systems at once through positive-sum compromise. There's significant variance in this recommendation, but because MIRI's work is so important, my current expected value for donating remains high.

Other organizations do important future-oriented research as well, and there are further charities in the effective-altruism movement that aim for similar goals at a more meta-level. I personally favor MIRI most because I think AI is the catastrophic risk where the interests of suffering reducers and happiness increasers most plausibly align. I'm more ambivalent about the Future of Humanity Institute (FHI) because its work on non-AI risks may (or may not) be detrimental for suffering reducers. Promoting memes favoring space colonization and especially disaster recovery seem quite harmful.

Past donations

In the remainder of this piece I'll describe where I used to donate. One reason my recommendations have changed recently is that Animal Ethics did not exist before 2013, and I had a full-time job that reduced how much bandwidth I had for direct research. However, I have also updated my views about the potential net impacts of the other causes as well.

Between 2009 and 2013, I gave $12K per year because my employer, Microsoft, matched employee contributions to US 501(c)(3) public charities dollar-for-dollar up to $12K. This opportunity for doubling was the main reason I donated each year rather than holding on to the money for better future opportunities.

  • Pre-2005: Before I realized that animal welfare mattered, I was (misguidedly, relative to my current values) an environmentalist and encouraged my parents to donate to environmental-preservation organizations like Natural Resources Defense Council.
  • 2005-2007: Soon after realizing that animals could suffer thanks to an article by Peter Singer, I came across a Vegan Outreach brochure. I visited the organization's website and was impressed by its apparent cost-effectiveness at creating vegetarians. I wrote a naive cost-effectiveness analysis of the organization and made small monthly donations.
  • 2008-2009: I discovered the Singularity Institute for Artificial Intelligence (SIAI, renamed to MIRI) in 2006. I was skeptical of the organization's relative neglect of animals and insufficient focus on suffering, but I thought its work was plausibly net beneficial in spite of that. In 2009, I donated to SIAI in an earmarked fashion to support a particular research topic that was more focused on suffering than what SIAI counterfactually would have done.
  • 2010-2011: I became worried that SIAI's focus on extinction-risk reduction might be net harmful, and I wasn't sure if a proverbial paperclip maximizer would be better for suffering reduction than a friendly AI. Concerned that SIAI donations might be net negative in expected value, I switched back to my previous choice, Vegan Outreach. While preventing direct factory-farmed suffering was motivating, the main benefit I saw was to spread the idea that animal suffering matters more generally, with ultimate spillover effects in terms of helping wild animals.
  • 2012: I met Nick Cooney of The Humane League and was impressed by his focus on measuring effectiveness. His results on the efficiency of veg ads were eye-catching, and I decided I wanted to donate toward veg ads. I asked Vegan Outreach about the idea, and they agreed to fund veg ads through their organization using my 2012 donation.
  • 2013: I donated to Animal Charity Evaluators (then called Effective Animal Activism) because it seemed to have higher expected leverage per dollar than direct donations to its object-level charity recommendations.
  • 2014-: Until things change, I plan to put most of my resources into my own foundational research on reducing suffering, and I may donate to Animal Ethics if we encounter funding shortfalls when trying to hire new employees. I still watch the other animal charities from the sidelines, as I do many of the other groups in the effective-altruism movement.

What do I think now of past charities?

Organization type environmental preservation veg outreacha MIRI (formerly SIAI)
Net sign Often probably net negative (e.g., habitat conservation), but sometimes probably net positive (e.g., opposing climate change). Positive with high variance. Very positive with high variance.
Good effects Conserving natural resources may improve prospects for future cooperation. Encourage the idea that animal suffering matters. Research on crucial considerations that can benefit suffering reducers.
Bad effects Preserve habitats and, along with them, lots of wild-animal suffering. Also, encourage memes of deep ecology, that nature is intrinsically valuable. Veg*ism may encourage support for wilderness preservation.b Much of the ethics promoted by MIRI is not focused on and is sometimes even antithetical to suffering reduction. That said, CEV is a step towards an optimal compromise morality, which is probably what we should ultimately shoot for with AI.
Unclear effects Sign of veg*ism for wild animals is unclear. Not obvious whether friendly AI is better or worse than paperclip maximization.

Estimating the overall impacts of a charity is really hard, and the uncertainty displayed in this table underscores the importance of foundational research before jumping to conclusions about specific activism projects.

  1. e.g., Vegan Outreach, The Humane League  (back)
  2. I now think this is more of a concern than I used to. When I meet new veg*ans outside of rationalist circles, they are always environmental preservationists, and they're usually more appalled by my views on wild-animal suffering than are non-veg*ans. Many veg*ans think wild-animal lives are net positive, even often those of insects. Now, there's some hope that this correlation is partly due to more general ideological trends (liberalism is associated with both environmentalism and support for animal welfare), and maybe marginal veg*ans are less likely environmentalist. Still, being veg*an will inevitably expose people to environmentalist friends and memes.  (back)