Maybe too much for a Draft Amnesty week, but I'd be excited for someone / some people to think about how we'd prioritise R&D efforts if/when R&D is ~automated by very powerful narrow or general AI. "EA for the post-AGI world" or something.
I wonder if the ITN framework can offer an additional perspective to the one outlined by Dario in Machines of Loving Grace. He uses Alzheimer’s as an example of a problem he thinks could be solved soon, but is that one of the most pressing problems that becomes very tractable post-AGI? How does that trade-off against e.g. increasing life expectancy by a few years for everyone? (Dario doesn't claim Alzheimer’s is the most pressing problem, and I would also be very happy if we could win the fight against Alzheimer’s).
I found this finding in the MCF 2024 survey interesting:
The average value to an organization of their most preferred over their second most preferred candidate, in a typical hiring round, was estimated to be $50,737 (junior hire) and $455,278 (senior hire).
This survey was hard and only given to a small number of people, so we shouldn't read too much into the specific numbers, but I think it's still a data point against putting significant weight on replacability concerns if you have a job offer for an org you consider impactful.
Survey respondents here (who all work at EA orgs like Open Phil, 80k, CEA, Giving What We Can) are saying that if they make someone a job offer, they would need to receive, in the typical case for junior staff, tens of thousands of dollars to be indifferent about that person taking the job instead of the next best candidate. As someone who's been involved in several hiring rounds, this sounds plausible to me.
If you get a job offer from an org you consider impactful, I suggest not putting significant weight on the idea that the next best candidate could also take the role and have just as much or more impact as you, unless you have a good reason to think you're in an atypical situation. There's often a (very) large gap!
FYI the question posed was:
Imagine a typical hiring round for a [junior/senior] position within your organization. How much financial compensation would you expect to need to receive to make you indifferent about hiring your second most preferred applicant, rather than your most preferred applicant?
(there's a debate to be had about how "EA org receiving X in financial compensation" compares to "value to the world in $ terms" or "value in EA-aligned donations" but I stand by the above bolded claim).
Full disclosure: I work at CEA and helped build the survey, so I'm somewhat incentivised to say this work was interesting and valuable.
Already so many EAs work at Anthropic that it is shielded from scrutiny within EA
What makes you think this? Zach's post is a clear counterexample here (though comments are friendlier to Anthropic) and I've heard of criticism of the RSPs (though I'm not watching closely).
Maybe you think there should be much more criticism?
"Why I'm long-term bullish on the Middle East" by Noah Smith has some great good news stories
- Wars in the Middle East are becoming less violent (contrary to many media narratives)
- The Middle East is building a ton of desalination plants to provide fresh drinking water
- Investment in solar energy is booming
"In other words, with solar power rising and oil becoming less important, and with its demographics in a favorable position, the Middle East is primed for an economic and political reinvention."
Good question! Yes, TL;DR large venues in major US/UK cities are more expensive per-attendee than smaller venues in other cities.
Eli covered this a bit in our last post about costs. There aren't that many venues big enough for EA Globals, and the venues that are big enough force you to use their in-house catering company, generally have a minimum mandatory spend, and significantly mark up the costs of their services. Our best guesses at why (from Eli's post):
- Big venues are just generally quite expensive to run (big properties, lots of staff, etc.).
- These venues are often empty, forcing them to charge more when they actually do host events.
- Catering costs are marked up in order to mark venue costs down. Many customers will anchor on an initial venue cost; by the time they hear the exorbitant catering fees later, they may feel it’s too late to switch. (We always ask to see both venue and catering costs up front.)
I suspect straightforward lack of competition also plays a role. As an extreme example, if there's only one venue in a city large enough for conferences and you want to run a conference there, they can basically charge what they want to.
Meanwhile, venues that can host 200–600 people (EAGx events) are easier to come by. EAGx organizers often secure university venues which are cheap but often more difficult to work with. Location does play a role, of course. You may not be surprised to learn that Mexico City, Bangalore and Berlin are cheaper than Oakland, London and Boston. But we also hosted events in Sydney and Copenhagen this year, so I think the above cost vs. size factor / availability of space plays a bigger role.
I do want to add that we are consistently impressed by EAGx and EA Summit organizers when it comes to resourcefulness and the LTR scores they generate given the lower CPA. The EA Brazil Summit team, for example, had food donated by the Brazilian Vegetarian Society. The bar for hustling in service of impact is continuously being raised, and we hustle on.
(Other team members or EAGx organizers should feel free to jump in here and push back / add more details.)
Flag that I didn't catch that this was an important announcement, and I think that's because it's posted by one user with initials. Hard to explicate exactly what's going on, but that made me think it was one anonymous user's reactions to an OP announcement rather than the real deal.
By contrast, the technical AIS RFP has three co-authors with full names, and I recognised them as people who work on that team. I'd guess posts with multiple full-name co-authors are more likely to be understood as important and therefore get more reach :)