Here are the biggest things I got wrong in my attempts at effective altruism over the last ~3 years.
1. I thought leafleting about factory farming was more effective than GiveWell top charities. I am now unsure about this. But I was way too confident based on evidence which I now think isn’t very good.
I probably made this mistake because of emotional bias. I was frustrated by people who advocated for global poverty charities for dumb reasons. A lot of them hadn’t thought much about animal suffering, which I thought was embarrasingly negligent. And a lot of them claimed to prefer human-targeting charities because of the stronger evidence base; I thought that if they really had that belief, they should either save their money just in case we found a great intervention for animals in the future, or donate it to the people who were trying to find effective animal right interventions.
I think that this latter argument was correct, but I didn’t make it exclusively. This mistake didn’t cost me much except $500 to Vegan Outreach which I kind of regret, and probably I looked dumb to the people who disagreed with me and noticed what mistake I was making.
(An alternative way of looking at this mistake is as a failure caused by tribalism–I believed correctly that animal welfare is a more promising cause than global poverty, and I identified pretty strongly with the animal-focused EAs over the global poverty EAs. And then I made the non-sequitur leap to “this particular animal-focused intervention is better than this particular global-poverty-focused intervention.”)
2. In 2014 and early 2015, I didn’t pay as much attention to OpenPhil as I should have. At the time I thought that even though Good Ventures has so much money that anything it wants to fund isn’t funding constrained, OpenPhil plausibly wouldn’t fund the causes I’m most tempted to fund–most obviously existential risk and farm animal welfare. This belief led me to focus much more intently on earning-to-give that was probably reasonable in hindsight.
Being wrong about OpenPhil’s values is forgivable, but what was really dumb is that I didn’t realize how incredibly important it was to my life plan that I understand OpenPhil’s values. I could have tried a lot harder to figure these things out earlier, but I didn’t think of that.
3. I wish I’d thought seriously about donating to MIRI sooner. In 2014 to mid 2015 I was agnostic about basic questions like “is MIRI doing anything useful”. These questions were hard to answer, and I didn’t really try until much later. (Honestly, I mostly didn’t think about this myself, I just made friends with people whose cause prioritization I trust, and let them convince me.)
Like my error #2, this is an example of failing to realize that when there’s an unknown which is extremely important to my plans but I’m very unsure about it and haven’t really seriously thought about it, I should probably try to learn more about it.