Very low confidence in this take, but I suspect that in university groups, Better Futures fieldbuilding looks much better under short timelines than technical safety research fieldbuilding. As far as I can tell, the barriers to doing important stuff in that field are much lower.
AI Safety is important - yes, but you don't have to pivot. Maybe you just don't have a horse in that race.
War makes me sad. The war in Russia-Ukraine. The war in Israel-Palestine. People impoverished and holding the butt of decisions they had no say in due to politics & power.
I really cannot do anything about this. I can vote for officials who care about peace and would advocate for that! But I cannot put Putin & Zelensky in the same room and make them:
I talk to individuals from various career paths who have fear about AI for various reasons and feel pressured to pivot, but I dont think it makes sense to advise everyone to simply pivot, especially if there is not a role for them to land.
If you're a nurse and enjoy nursing, please stay and help your patients. If you're a teacher and enjoy teaching, please stay and enrich the lives of children. If you're a PhD in pharmaceutics, please stay and help the creation of medicine to address ailments.
Some people have recently been talking about a decrease in activity on the forum. I've written before about the possibility of a yearly theme. I don't know if it would work, but it seems like it might be worth a shot to counter the risk of stasis. It's somewhat hard to articulate why I think things like this might be valuable, but one of the barriers that a community can face in terms of moving forward is that there's so many potential different topics that people could talk about. Purely relying on the marketplace of ideas often means that it's hard to build momentum around any particular direction the community could be exploring. Now maybe running weeks focused on particular topics already covers this and does it better, but I think it's at least worth exploring.
It’s been a busy week interviewing hiring managers at AI Safety organizations for my SPAR project, to fully understand the bottlenecks that affected both hiring teams and candidates.
I published a Playbook v0.1 of my learnings last week (would love some feedback!) , as I continue to work towards publishing a final Playbook to help hiring teams to improve their hiring processes.
Additionally, I am so appreciative for all the warm introductions that I’m receiving - this community really supports each other.
These hiring bottlenecks continue to be:
Here’s some interesting things I learned in my calls with hiring managers this week:
You are assuming excruciating pain is 45.6 (= 11.17/0.245) times as intense as hurtful pain. This implies 1 min of “severe burning in large areas of the body, dismemberment, or extreme torture” (excruciating pain) is as bad as 45.6 min of “awareness of Pain is likely to be present most of the time” (hurtful pain). I believe 1 min of excruciating pain is way worse.
Excruciating pain in humans would “mark the threshold of Pain under which many people choose to take their lives rather than endure the Pain”. I doubt people taking their lives to avoid 1 min of excruciating pain would do so to avoid 45.6 min of hurtful pain. In this pain, “Individuals can still conduct routine activities that are important in the short-term (e.g. eating, foraging) and perform cognitively demanding tasks, but an impairment in their ability or motivation to do so is likely to be observed”.
I made a similar comment in 2025, and this year privately. I know you have considered them. However, I do not understand why you believe the intensity of excruciating pain is so low relative to what I consider plausible.
At this point, it feels like the bottleneck being "talent" vs. "context" or anything else is just semantics.
Talent that has years of experience but doesn't have the specific knowledge I want is not the talent I need. I need a different talent. Therefore, it's still a talent bottleneck.
Maybe it would be better if AI safety orgs could be extremely more exact about what talent means to them, but that also leaves the field fairly open to a type of gaming a lot of people seem to want to avoid.
The field does act like a false positive hire is immensely worse than a false negative, though, which might not be the case.
I don't have a point, these are just random observations.
Thanks for writing this! I agree it is a bit odd that both 1) funders say they are short on for grantakers, and 2) lots of people would love to do AI safety grantmaking. I basically think funders are right to keep a high bar, and have strong AI safety knowledge/experience as an important hiring desideratum. I think the more logistical/operational aspects of investigating and making grants can be learned quite quickly on the job, whereas developing taste and judgement about AI strategy and safety is harder. So I think the current strategy of taking existing research-y/policy/technical AI safety people and training them up to be grantmakers seems better than taking existing non-AI grantmakers and training them up in AI safety (though of course running career transition programs for people getting into AI safety is also valuable!).
(I am biased by being one of those EA/AI safety research-y people who transitioned into grantmaking.)
If we want EA to be a decentralised movement, it’s worth looking at how other movements/structures move money from the centre to national, regional, and local orgs. Here are a few comparisons.
In the big federations, money flows up. Amnesty’s national sections sent just over a third of what they raised to the international budget in 2013. Greenpeace works in a similar way. Because the sections fund the centre, they also have a formal say in how it’s run.
EA works the other way round. Most of the movement building money comes from one funder, Coefficient Giving, and passes through CEA. Three examples have a similar shape (it's hard to find something perfectly analogous to EA):
At CEA, operational spending in 2025 was $21.8m, Claude tells me that Community Building Grant costs fell 22% that year. I don’t know what share of CEA’s budget goes to local and national orgs, and I couldn’t find a published figure. In 2026 the CBG portfolio covers 11 orgs and is closed to new applicants.