I lead a small think tank dedicated to accelerating the pace of scientific advancement by improving the conditions of science funding. As well, I’m a senior advisor to the Social Science Research Council. Prior to these roles, I spent some 9 years at Arnold Ventures (formerly the Laura and John Arnold Foundation) as VP of Research.
Stuart Buck
Karma: 1,244
I don’t think Moravec’s paradox is a real paradox in the first place. The reason we think of activities like walking, using your hands, being able to speak language, etc. as easy is because we don’t have any long-term memories from the first few years of life when we had to figure all of that out from scratch. https://x.com/stuartbuck1/status/1798547161489231928
The main problem is that even if you do this as an individual, the whole setup creates a new (and mostly bad) set of social norms about what everyone else is doing, whether you can even say hi to them, etc.
“One-on-one (1:1) meetings are widely recognized as one of the most valuable aspects of EA Global (EAG) and EAGx conferences”
Speaking just for myself, the shift to endless one-on-one meetings is the single biggest reason I have no desire ever to attend another EA conference. [I’ve been to the following: EA Global 2015 at Google (no Swapcard or one-on-ones there), EA Global in DC 2022, and EA Global Feb. 2024]
There are pros and cons.
Pros:
More social and technological affordances for meeting up with new people
Cons: You list lots of ways one-on-ones can go wrong, but there are many more.
There is high variance in whether conversations are useful/pleasant. There are lots of people that I’d like to talk to for 2-5 minutes just to meet them and assess whether we have enough in common to continue further. But if they’ve scheduled a 25-minute 1-on-1 with me, it feels really awkward to end things after 2 minutes. And the same is true vice versa, I would have to assume.
Speaking of brand-new forms of social awkwardness, one-on-one meetings manage to create yet another one. In any normal social setting, if you see an old friend casually chatting with someone, you can go up and say “hi” and then be introduced to their conversational partner and it turns into a group conversation. In a setting dominated by scheduled one-on-ones, going up and saying “hi” to an old friend can be met with a stern, “Sorry, we’re having a one-on-one.” Ridiculous.
The dominance of one-on-ones takes people away from enjoying common experiences that would be much more valuable for 95% of them than any random one-on-one. At the last EA conference I went to, Patrick Collison and Alexander Berger had an onstage debate of sorts about EA. At a normal conference, this would be a standing-room-only event, and everyone afterwards would have a shared experience to talk about. [And, while they weren’t obvious about it, Dustin Moskovitz, Cari Tuna, and Joseph Gordon-Levitt were all in the crowd.] At this conference, the room wasn’t even full, because the vast majority of attendees were off having some one-on-one that was almost guaranteed to be more uninteresting than hearing Patrick Collison’s thoughts. What? Dustin Moskovitz thinks it is worth his time to spend an hour listening to a fascinating conversation, but you’re off having a one-on-one with some random person who you likely won’t even remember the next day?
Non-stop one-on-one conversations are way more mentally exhausting than any other normal conference I’ve ever been to. Part of this is connected to Con #1 above—some people are difficult to talk to for the full time. That could be the fault of either of us, to be sure, but it is nonetheless tiring to have to carry the conversation for the whole time. At any normal event, you can just make a polite excuse after a minute, “Sorry, I need to use the restroom,” or “hey, I just spotted someone across the room,” or whatever.
For me, the net value is overwhelmingly negative.
Always interesting to read a piece by someone who is actually familiar with science, as opposed to science fiction from folks who are afraid of diamondoid bacteria or whatever else they have imagined.
There are amazing journalists being laid off every year from jobs that pay less than 1/4th that much.
I thought this piece was going to be about the firm Anthropic . . . anyway, interesting subject, carry on!
OK, fair enough, what I said was perhaps a bit overstated. It is even more overstated to refer to a “conference filled with racist speakers” etc.
The main debate here is whether people who ever aid controversial things should be allowed to attend an event at all, and/or to give a talk about unrelated issues.
Ah, that is a fair point!
Great comment and overview of the event, which I very much enjoyed.
Was anyone there who had ever uttered a previous phrase or sentence with which I might disgree, even firmly so? Almost certainly.
I mean, Eliezer was there, and he has suggested that human infants might be susceptible to killing up to 18 months (https://x.com/antoniogm/status/1632162012229693440), which I regard as unbelievably monstrous.
But even if someone said something monstrous, I’m still willing to hear them out, to attend a conference with them, and to attempt to persuade them otherwise (if it comes up). And who knows, maybe some belief of mine might turn out to seem monstrous to other people. I should hope they’d try to engage with me.
Trying to cancel folks because they spoke at an event but another speaker said a bad thing 15 years ago—that’s an absurd level of guilt by association.
Is the consensus currently that the investment in Twitter has paid off or is ever likely to do so?
I could imagine making that case, but what’s the point of all the Givewell-style analysis of evidence, or all the detailed attempts to predict and value the future, if in the end, what would have been the single biggest allocation of EA funds for all time was being proposed based on vibes?
I did think Harris could have been slightly more aggressive in his questioning (as in, some level above zero). E.g., why would MacAskill even suggest that SBF might have have been altruistic in his motivations, even though we now know about the profligate and indulgent lifestyle that SBF led? MacAskill had to have known about that behavior at the time (why didn’t it make him suspicious?).
And why was MacAskill trying to ingratiate himself with Elon Musk so that SBF could put several billion dollars (not even his in the first place) towards buying Twitter? Contributing towards Musk’s purchase of Twitter was the best EA use of several billion dollars? That was going to save more lives than any other philanthropic opportunity? Based on what analysis?
This statement cracked me up for some reason: “At this point, we are not actively making grants to further investigate these questions. It is possible we may do so in the future, though, so if you plan to research any of these, please email us.”
I.e., this isn’t an RFP (request for proposals). Instead, it’s more like a RFINP, BMITFWDEK? Request for Information Not Proposals, But Maybe In The Future (We Don’t Even Know)?
OK, mostly joking—in all seriousness, I haven’t seen wealthy philanthropies release lists of ideas that are hopefully funded elsewhere, but maybe that actually makes sense! No philanthropy can fund everything in the world that might be interesting/useful. So maybe all philanthropies should release lists of “things we’re not funding but that we hope to see or learn from.”
I don’t have personal references, but a ton of great journalists have gotten laid off in 2023, and they never were paid that much in the first place (not being TV or celebrity journalists).
https://www.poynter.org/business-work/2023/buzzfeed-news-closed-180-staffers-laid-off/
https://www.sfgate.com/tech/article/wired-layoffs-conde-nast-magazine-18550381.php
https://www.washingtonpost.com/style/media/2023/10/10/washington-post-staff-buyouts/
Great journalists are getting laid off all the time these days. You could find any number of professional and highly accomplished journalists for a tiny fraction of $800k per year.
$800k per year? For one person to do investigative journalism? What would all that money be spent on?
“Completely remove efficacy testing requirements, making the FDA a non-binding consumer protection and labeling agency.”
That seems to be how the supplement market works under DSHEA. Is there any evidence that stuff marketed as supplements is any better at curing disease, preventing disease, improving health, etc., than the FDA pathway? Indeed, is there any evidence that the supplement market works well at all? We’ve known for decades that antioxidants don’t actually improve health, yet they are still a huge market. https://pubmed.ncbi.nlm.nih.gov/17327526/
So let me put it this way:
If there is a future bioterrorist attack involving, say, smallpox, we can disaggregate quite a few elements in the causal chain leading up to that:
The NIH published the entire genetic sequence of smallpox for the world to see.
Google indexed that webpage and made it trivially easy to find.
Thanks to electricity and internet providers, folks can use Google.
They now need access to a laboratory and all the right equipment.
Either they need to have enough resources to create their own laboratory from scratch, or else they need to access someone’s lab (in which case they run a significant risk of being discovered).
They need a huge amount of tacit knowledge in order to able to actually use the lab—knowledge that simply can’t be captured in text or replicated from text (no matter how detailed). Someone has to give them a ton of hands-on training.
An LLM could theoretically speed up the process by giving them a detailed step-by-step set of instructions.
They are therefore able to actually engineer smallpox in the real world (not just generate a set of textual instructions).
The question for me is: How much of the outcome here depends on 6 as the key element, without which the end outcome wouldn’t occur?
Maybe a future LLM would provide a useful step 6, but anyone other than a pre-existing expert would always fail at step 4 or 5. Alternatively, maybe all the other steps let someone let someone do this in reality, and an accurate and complete LLM (in the future) would just make it 1% faster.
I don’t think the current study sheds any light whatsoever on those questions (it has no control group, and it has no step at which subjects are asked to do anything in the real world).
Not sure if this is responsive, but while most people (barring disability) do reach a decent level of proficiency at navigating the world, I think the variance can still be huge. Look at the feats of Olympic gymnasts or professional basketball players—hardly anyone could reach that level no matter how much they trained.