At some point, the statistics can almost become self-fulfilling. Many (if not most) people prefer to be in spaces where a decent fraction of the people are like them, where they feel like people like them belong. If there are few people like them, they may get the impression that this isn’t a space for them. “People like them” could refer to race, religion, politics, socioeconomic status, or any number of identity characteristics. Thus, at some point, I would expect to see lopsided statistics in categories like race, religion, politics, or SES mostly continue even if there was zero discrimination and the original causes for those lopsided statistics were no longer relevant. In other words, I wouldn’t expect the statistics to change that much in those categories without taking special steps to improve representation.
For instance, I don’t think the width of one’s moral circle would significantly predict interest in AI safety since AI doom will just as equally harm one’s “family, tribe, and nation.” And it’s plausible that conservatives (especially religious conservatives) would be overall more skeptical of AI, and big tech companies, than average.
A few other points:
I wonder if conservatives feel less welcome in EA social spaces than they would in EA professional spaces, which would make this one of the places where the porous boundaries between EA as social group and EA as professional group could negatively affect impact. (I identify as a centrist, so am not in a position to answer that.)
I’d also be curious how often the people who responded as right, center-right, or even center display their right-leaning beliefs in EA spaces, versus how frequently those who responded as left or even center-left do. My guess is that a reasonable observer would think EA is even more politically concentrated on the left than the poll results would suggest. Same question for religion.
Incidentally, I do think that being coded too far on the left is an impediment to maximum impact. In addition to recruitment/retention issues with centrists and right-of-center people, I think some people in EA are prone to underestimate the importance of political action as a strategy to get certain things done . . . and at least in the US, actually getting stuff done can be easier if you’re not coded too strongly on one side of the ideological spectrum.
At some point, the statistics can almost become self-fulfilling. Many (if not most) people prefer to be in spaces where a decent fraction of the people are like them, where they feel like people like them belong. If there are few people like them, they may get the impression that this isn’t a space for them. “People like them” could refer to race, religion, politics, socioeconomic status, or any number of identity characteristics. Thus, at some point, I would expect to see lopsided statistics in categories like race, religion, politics, or SES mostly continue even if there was zero discrimination and the original causes for those lopsided statistics were no longer relevant. In other words, I wouldn’t expect the statistics to change that much in those categories without taking special steps to improve representation.
For instance, I don’t think the width of one’s moral circle would significantly predict interest in AI safety since AI doom will just as equally harm one’s “family, tribe, and nation.” And it’s plausible that conservatives (especially religious conservatives) would be overall more skeptical of AI, and big tech companies, than average.
A few other points:
I wonder if conservatives feel less welcome in EA social spaces than they would in EA professional spaces, which would make this one of the places where the porous boundaries between EA as social group and EA as professional group could negatively affect impact. (I identify as a centrist, so am not in a position to answer that.)
I’d also be curious how often the people who responded as right, center-right, or even center display their right-leaning beliefs in EA spaces, versus how frequently those who responded as left or even center-left do. My guess is that a reasonable observer would think EA is even more politically concentrated on the left than the poll results would suggest. Same question for religion.
Incidentally, I do think that being coded too far on the left is an impediment to maximum impact. In addition to recruitment/retention issues with centrists and right-of-center people, I think some people in EA are prone to underestimate the importance of political action as a strategy to get certain things done . . . and at least in the US, actually getting stuff done can be easier if you’re not coded too strongly on one side of the ideological spectrum.