1.) Amount of money going to meta-stuff vs. amount of money going to object-level stuff. What is the total budget of CEA + GWWC + 80K + Charity Science + TLYCS + ACE + …? What is the total amount of money being moved to GiveWell top charities + MIRI + ACE top charities + …?
2.) Average meta-level of meta-stuff. Is a typical project one level above (e.g., Charity Science fundraising for GiveWell top charities) or 2+ levels above (e.g., CEA funding a team to assess the EA movement for gaps in meta-orgs)? What is the level weighted by total funding per project?
We can measure this. Measuring (1) seems like it be easier than measuring (2). For (1), we just need to check the funds meta-charities have received, vs. the proportion of the funds direct work receives that is from effective altruism. Measuring (2) seems like it would require contacting and conversing with the executivies of meta-charities. That’s easy enough on it’s own, except they’re probably busy, and it would take them time to figure out exactly how meta each of their projects are, time they might rather spend running their organization. It’d be easier to get them to respond if we informed meta-charities that, e.g., providing this information would be incorporated into a global transparency report on meta-charity on the EA Forum, that we expect being included in the report would boost the meta-charity’s prospects, and that their conspicuous absence or lack of information provided might just raise more questions or skepticism of them in the future.
Anyway, I’d be willing to do all this to gauge if effective altruism is indeed falling into a “meta trap”. Let me know what you think.
What is the level weighted by total funding per project?
I don’t understand what this means. Can you put it another way? Feel free to use real or hypothetical examples.
What is the level weighted by total funding per project?
I don’t understand what this means. Can you put it another way? Feel free to use real or hypothetical examples.
For example, imagine there are only two projects in the entire EA movement. Christie’s Chapters is an EA group that supports volunteers in creating local chapters, which is three levels removed from direct impact. Boris’s Bednets is a fundraising org that raises money for AMF, which is one level removed from impact. The average meta level is thus (3 + 1)/2 = 2.
Another way to look at it is to weigh by funding. If Christie’s Chapters has $1M but Boris’s Bednets has $10M, we should weigh Boris’s Bednets much more highly since more resources are going to it. We thus apply a weighted average (3$1M + 1$10M) / $11M = 1.18.
What sort of feedback signals would we get if EA was currently falling into a meta-trap? What is the current state of those signals?
Two ideas might be:
1.) Amount of money going to meta-stuff vs. amount of money going to object-level stuff. What is the total budget of CEA + GWWC + 80K + Charity Science + TLYCS + ACE + …? What is the total amount of money being moved to GiveWell top charities + MIRI + ACE top charities + …?
2.) Average meta-level of meta-stuff. Is a typical project one level above (e.g., Charity Science fundraising for GiveWell top charities) or 2+ levels above (e.g., CEA funding a team to assess the EA movement for gaps in meta-orgs)? What is the level weighted by total funding per project?
William_S:
We can measure this. Measuring (1) seems like it be easier than measuring (2). For (1), we just need to check the funds meta-charities have received, vs. the proportion of the funds direct work receives that is from effective altruism. Measuring (2) seems like it would require contacting and conversing with the executivies of meta-charities. That’s easy enough on it’s own, except they’re probably busy, and it would take them time to figure out exactly how meta each of their projects are, time they might rather spend running their organization. It’d be easier to get them to respond if we informed meta-charities that, e.g., providing this information would be incorporated into a global transparency report on meta-charity on the EA Forum, that we expect being included in the report would boost the meta-charity’s prospects, and that their conspicuous absence or lack of information provided might just raise more questions or skepticism of them in the future.
Anyway, I’d be willing to do all this to gauge if effective altruism is indeed falling into a “meta trap”. Let me know what you think.
I don’t understand what this means. Can you put it another way? Feel free to use real or hypothetical examples.
For example, imagine there are only two projects in the entire EA movement. Christie’s Chapters is an EA group that supports volunteers in creating local chapters, which is three levels removed from direct impact. Boris’s Bednets is a fundraising org that raises money for AMF, which is one level removed from impact. The average meta level is thus (3 + 1)/2 = 2.
Another way to look at it is to weigh by funding. If Christie’s Chapters has $1M but Boris’s Bednets has $10M, we should weigh Boris’s Bednets much more highly since more resources are going to it. We thus apply a weighted average (3$1M + 1$10M) / $11M = 1.18.
This makes lots of sense, thanks. I can now carry on with the project of checking all this money moved, including with weighted averages.
Was this selection inspired by Boris Yakubchik?
lots of excitement, little in the way of new or surprising successes.