I’m interested in AI strategy, AI governance, global priorities research, and forecasting. I am currently learning about machine learning, formal epistemology (particularly interested in the Sleeping Beauty Problem), and statistical techniques for forecasting.
10xRational
Karma: 43
Hi Mo, thanks for the comment. I have the sense that participating in multiple activities (i.e. λ>1) enables community-builders to connect between different individuals, organizations, and resources as they act as central nodes connecting individuals to resources. I agree that smaller/younger groups would likely have a lower λ per your intuition above. The model above does not capture this intuition, but the power of the individual community builder’s network is also quite important for determining the extent of impact. I have the sense that community builders with a larger network should spread their focus more widely to maximize the chances for cross-pollination, while community-builders in more localized contexts may find it more optimal to focus on one particular activity to avoid getting distracted. Whether trying to increase λ (at the expense of perhaps spending less time on bridging out into other activities) is therefore dependent on the community-builder’s existing network and capacity; full-time and well-connected community-builders should probably care about increasing λ more than part-time or localized community-builders, who probably benefit from focussing very particularly on one particular programme or intervention that helps build the EA community.