Thanks for expanding! I appreciate the distinction between "language" and "concepts" as well as your thoughts on using language for in-group signaling and too much in-group hiring.

To follow up on this:

it's not important to me that people know a lot about in-group language, people, or events around AI safety

I can see that people and events are less important, but as far as concepts go, I presume it would be important for them to know at least some of the terms, such as x/s risk, moral patienthood, recursive self-improvement, take-off speed, etc.

As far as I know, really none of these are widely known outside of the AIS community, or do you mean something else by in-group language?

Ah, right. x) Thanks so much for pointing this out!

Yeah I agree about this case, I will actually take it out!

Thanks for sharing this, David!

thanks for flagging, I fixed it now!

Sorry for the late reply, Vasco. Thanks for your comment! I echo some of Chris's points, but wanted to add some scattered thoughts:

I think the cost-effectiveness of AIS groups vs Bluedot would definitely depend on a given group, as I expect there to be large differences between different groups. The best ones might be more cost-effective than Bluedot, but I doubt anyone looked at this very rigorously (this analysis is great, but doesn't include a Bluedot-like program). If you have the money, you likely want to fund both even if one is slightly better than the other.

 The volunteer-run initiatives are likely more cost-effective kind of by default (though see next point).

Another consideration is a tradeoff between impact and cost-effectiveness. E.g if Bluedot is not funding constrained, then it might make sense for them to optimize for the largest impact, even if that comes with the price of slightly decreased cost-effectiveness. An example of this could be diminishing returns on spending money on LinkedIn ads. This could mean that the marginal impact of money donated to them is smaller, but funders still might prefer this to having to spend time on evaluating 10 AIS groups due to time costs)

Great post, thanks for sharing!

Furthermore, it is probably a non-sum zero game, and more efforts on MA might not come at the expense of AI safety money.

Agreed. As far as I know, Polaris Ventures is interested in the s-risk space but not fund more "traditional" AI Safety work.

Hey Egg, thanks for your comment! Here are my thoughts:

but far more often than that I point new people to the BlueDot curriculum. I commonly see others doing the same; I think it's become the default AIS 101 reading. Maybe you're mistaking that for people pushing the BlueDot course on everyone new to the field?

This totally makes sense, I do the same, though I think if people have the opportunity to take a "live" course that is more beneficial. What this post aims to respond to is the notion that, given that Bluedot exists as an organisation, people conclude that there is no need to start local fieldbuilding initiatives (something I come across quite often). Hope that clarifies!

AI safety (other than governance) isn't at all a local problem, and so there's no particular reason to focus on local groups.

Agreed! However, looking at the many benefits that such initiatives provide (some of which you mentioned, and the others I outline in the post) I think it is justified to run them.

[on AIS being management constrained] That's not obvious to me; I do think there are constraints there but my sense is that the field is currently mainly bottlenecked by funding (1, 2)

I could concede that the main bottleneck is funding right now. My current guess on funding gaps is that up until now, it was possible to get a small "moonshot" grant from LTFF relatively easily (this might change now that they pivoted to doing funding rounds), but then projects will fail to maintain funding once they need over 100k USD. For orgs that can fundraise from OP, money is less of an issue.

Why are they more likely to give AIS the benefit of the doubt? Won't that be most likely to happen if their exposure is to the highest-quality course they have access to?

What I mean here is that if you are introduced to a local AIS community through a friend who is also part of that group, you are more likely to give them the benefit of the doubt even if the course is not run as professionally as Bluedot's. Compared to such a person, I expect it's better for an experienced professional to take Bluedot's course instead of one organised by university students or fresh graduates. The quality of materials is important in either case!