2 days left to apply to mentor for Pivotal's Q1 2027 AI Safety Fellowship (January–April 2027, 15 weeks).
I, quite self-servingly, believe that most researchers in the field should spend 5-10% of their time supporting junior researchers. Mentorship is the bottleneck right now to get great talent into the field. Pivotal has a large pool of exceptionally great junior researchers to choose from.
We've changed the format in two ways. 1. The fellowship now runs 15 weeks instead of 9. 2. We also no longer offer extensions by default (though we may still do this for few exceptional projects). Instead, we aim to publish the research quickly and place fellows at orgs as soon as possible. We also will arrange seed funding for strong projects that should become new orgs.
If you know anyone who might be a good fit, forward it to them. If you have questions, please reach out!
It’s strange because Caroline’s message (especially the linked doc) implies that she believes she acted unethically and illegally. It would be weird for Austin to hire Caroline while disagreeing about this.
Thanks Sarah for the thoughtful response! I agree that women's suffrage is an interesting case to consider.
But it neither fully eliminates that root cause, nor does it fully eliminate the specific problem (women's situation improved after being granted the vote, but it didn't immediately switch to "no more disadvantageous outcomes for women in society"); nor can I now come up with a more comprehensive systemic intervention that would achieve these outcomes in this example.
This is a good point. Of course, an intervention can do a huge amount of good without fully eliminating the root cause or the specific problem.
Part of what prompted my comment was that I got the impression that the claim was that people working on systemic change are interested in permanent solutions, whereas many EA interventions merely fight symptoms. I don’t think you explicitly claim this, on a reread. I'd probably disagree with this framing :)
For example, an organisation fighting malaria might identify the lack of a fair distribution of resources to afford medicine and protective equipment as one root cause of the problem.
The most plausible, to me, root cause solution for this problem is a gene drive, which is not systemic change.
The biggest historical example of a huge problem being solved via root cause is likely smallpox eradication, which was a technical intervention. (It of course also required a a lot of systems-level work.)
Do you have an example of a huge problem that's been solved at its root cause through systemic change? Best I can think of is the Montreal Protocol. But may not count according your definition of “remov[ing] the problem indefinitely”, because the success depends on the protocol remaining in place.
I see a tension between addressing root causes (needs to be targeted on a very specific pathway – virus, chemical, mosquito, etc.) and your account of systemic change (needs to treat a problem "at large" and should look "beyond a narrow frame").
you are supposed to hedge your positions in order to prevent an unexpected situation from tanking the whole fund
It's my understanding that this hasn't been the definition of a hedge fund for some time (ChatGPT agrees). I'm assume that the investors knew that Situational Awareness was not a hedge fund in the sense you're describing one.
Of course, it might still be possible that investors weren't aware of the amount of risks the fund was taking on.
I found the 'founder deployment by end-2026' the hardest to set. It comes a bit as a surprise at the end, as I was already taking into account some considerations before, and the descriptions seem to do as well (e.g. "assets after lockups, taxes, sale timing", and "execution delays").
I'm really excited about you & Sarah starting this! :)
I wouldn’t invest community building efforts in different settings until I was confident this ecosystem was thoroughly supported
It's pretty notable that you're only focusing on North American universities. Do you think this subset of universities is your comparative advantage? Or do you think that most community builders should primarily focus on that subset?
If it's the latter, I'd be interested in hearing your thoughts on why efforts into other top universities (Oxford, Cambridge, Imperial, UCL, ETHZ, NUS, TUM, Melbourne, etc.) should be lower priority.
The first paragraph acknowledges this, and I don't think it's accurate to describe Tobias as dunking on vegans here?
2 days left to apply to mentor for Pivotal's Q1 2027 AI Safety Fellowship (January–April 2027, 15 weeks).
I, quite self-servingly, believe that most researchers in the field should spend 5-10% of their time supporting junior researchers. Mentorship is the bottleneck right now to get great talent into the field. Pivotal has a large pool of exceptionally great junior researchers to choose from.
You can find more about why you might want to mentor & what it involves here: pivotal-https://research.org/mentoring-at-pivotal
We've changed the format in two ways.
1. The fellowship now runs 15 weeks instead of 9.
2. We also no longer offer extensions by default (though we may still do this for few exceptional projects). Instead, we aim to publish the research quickly and place fellows at orgs as soon as possible. We also will arrange seed funding for strong projects that should become new orgs.
If you know anyone who might be a good fit, forward it to them. If you have questions, please reach out!
It’s strange because Caroline’s message (especially the linked doc) implies that she believes she acted unethically and illegally. It would be weird for Austin to hire Caroline while disagreeing about this.
https://substack.com/@yearningslav/p-213445133
I got "0 conflicts, 1 bullet bitten", with the bullet being the repugnant conclusion.
Does this mean that you have 6 bullets you'd prefer over having to bite the repugnant conclusion? :)
Thanks Sarah for the thoughtful response! I agree that women's suffrage is an interesting case to consider.
This is a good point. Of course, an intervention can do a huge amount of good without fully eliminating the root cause or the specific problem.
Part of what prompted my comment was that I got the impression that the claim was that people working on systemic change are interested in permanent solutions, whereas many EA interventions merely fight symptoms. I don’t think you explicitly claim this, on a reread. I'd probably disagree with this framing :)
The most plausible, to me, root cause solution for this problem is a gene drive, which is not systemic change.
The biggest historical example of a huge problem being solved via root cause is likely smallpox eradication, which was a technical intervention. (It of course also required a a lot of systems-level work.)
Do you have an example of a huge problem that's been solved at its root cause through systemic change? Best I can think of is the Montreal Protocol. But may not count according your definition of “remov[ing] the problem indefinitely”, because the success depends on the protocol remaining in place.
I see a tension between addressing root causes (needs to be targeted on a very specific pathway – virus, chemical, mosquito, etc.) and your account of systemic change (needs to treat a problem "at large" and should look "beyond a narrow frame").
Interested to hear your thoughts :)
It's my understanding that this hasn't been the definition of a hedge fund for some time (ChatGPT agrees). I'm assume that the investors knew that Situational Awareness was not a hedge fund in the sense you're describing one.
Of course, it might still be possible that investors weren't aware of the amount of risks the fund was taking on.
I found the 'founder deployment by end-2026' the hardest to set. It comes a bit as a surprise at the end, as I was already taking into account some considerations before, and the descriptions seem to do as well (e.g. "assets after lockups, taxes, sale timing", and "execution delays").
I'm really excited about you & Sarah starting this! :)
It's pretty notable that you're only focusing on North American universities. Do you think this subset of universities is your comparative advantage? Or do you think that most community builders should primarily focus on that subset?
If it's the latter, I'd be interested in hearing your thoughts on why efforts into other top universities (Oxford, Cambridge, Imperial, UCL, ETHZ, NUS, TUM, Melbourne, etc.) should be lower priority.