Thanks!

Obviously these results (and last year's) spark a concern about diversity. Has the EA community made any attempts to analyze, understand, and seek to remedy the causes behind the lack of gender and racial diversity based on the results of these surveys?

While most communities grow within a defined demographic at first, advocacy and community-building is a main goal and tenet of effective altruism. I'm concerned that numbers this skewed have a negative effect on this goal, and would love to know if the community has sought to understand and address this yet.

Thank you for sharing! This type of critical feedback is much needed in the EA community (for many of the reasons you stated).

"A potential danger with an EA start-up is that you might, as I did, find a hypothetical solution and try to retro-fit a business case." This is a really important takeaway from your conclusion, and I'm glad to hear it mentioned. A good piece of business/entrepreneurship advice I'd heard thrown around in startup circles before is to fall in love with the problem, not the solution. Keeping a constant internal check to evaluate how a venture, whether it be a new startup or a non-profit working on a global health intervention, is tackling the problem as effectively as possible ensures that venture is on the right track and producing the right solutions. Optimizing for the solution itself can veer those ventures off course.

Finally, I'd like to point out that you did find success here, even if the mobile application itself was a failure. By starting work on the PhD and the app at the same time, you gave yourself much more career capital than if you'd worked on one alone. In this case, the capital was not transferrable across both sides, but there's an alternate universe that exists where the work on your app led to a industry career. You kept your options open, and that's led you to a place where you're more confident in the career you've chosen. If other EAs find themselves in a similar opportunity (i.e. working on a side business and academia at the same time), it's probably a good thing to gain experience in both for awhile. The incremental amount of improvements each individual project would see if you went at it full time (i.e. one more published research paper, a fully functioning machine learning feature) likely do not matter in the long term. If you take value from the experience, and not the outcome, I think there are many positives to be had from this kind of work.