Thanks for clarifying. Using Scott's analogy, I'd respond by pointing to

In this case a yoga class might still benefit by making it super-clear that men are welcome and removing a couple of things that might make men uncomfortable

At present, going by the survey results, 9.8% of LessWrongers identify as female. (And 9.9% as women.) Quoting Wikipedia:

Women’s representation in the computing and information technology workforce has been falling from a peak of 38% in the mid-1980s. From 1993 through 1999, NSF’s SESTAT reported that the percentage of women working as computer/information scientists (including those who hold a bachelor’s degree or higher in an S&E field or have a bachelor’s degree or higher and are working in an S&E field) declined slightly from 33.1% to 29.6% percent while the absolute numbers increased from 170,500 to 185,000. Numbers from the Bureau of Labor Statistics and Catalyst in 2006 indicated that women comprise 27-29% of the computing workforce. A National Public Radio report in 2013 stated that about 20% of all US computer programmers are female.

I don't think either hypothesis ('women are socialized to be less interested in computer science' and 'women interested in computers get driven out by differential treatment by computer science authorities and communities') predicts that we'd be doing worse at gender representativeness over time. We'd expect both causes for inequality to be lessening over time, as society becomes more progressive / feminist / egalitarian. It is clear, however, that something we're doing is responsible for the rarity of women in such communities, and that this something can shift fairly rapidly from decade to decade. So, whatever the mechanism is, it looks plausibly susceptible to interventions.

If we grant that LessWrong has the power to improve its gender ratio without degrading the quality of discussion, then the only question is whether we prefer to retain a less diverse community. And it would be surprising to me if we have no power to move things in that direction. If we became merely as welcoming as computer science in general is today, we'd double the proportion of women at LessWrong. from 10% to 20%; if we became as attractive as computing and IT were in the '80s, or as economics is today, we'd rise to 30% or 40%; and if we had proportionally as many women as there are in psychology today, we'd be up to 70% women and have the opposite problem!

When we're doing worse than the worst of the large fields that can be claimed to have seeded LW, it's probably time to think seriously about solutions. (And, no, 'hey what if MIRI started a Pinterest account' does not qualify as 'thinking seriously about gender inclusivity'.)

Overall, I agree with Ben Kuhn's points on this issue.