That’s partially true I think. However, some EA orgs aren’t funding constrainted at all, therefore they hire people that’s better than a certain bar, not hiring people in a limited number. In this, you get whole credit, because even if you decide not to work there, there won’t be another people hired
jackchang110
Karma: 336
[Question] How useful is hardware knowledge in AI s-risks research?
Thanks for your willingness to type down your critique. Your idea is basically: Maybe I can’t become a PR 90 researcher, but at least I probably could become PR 90 at something and collaborate with others and make impact.
But my critique is : Suppose you’re PR 90 at writing skills but average at every other things. Maybe you can apply for writing position in EA world, to help researchers publish better articles/papers. But it’s hard to get in EA world, if you work in non-EA world, it seems impossible to reduce AI s-risks if you’re only good at writing. You’ll still get a writing job in non-EA world, but what you’re going to write is probably not related to AI s-risks at all. It seems only people with the skill of research or policy can make impact for AI s-risks in the non-EA world(such as implementing safety designs). There’s in fact not that many skills that can make influences in the non-EA world, especially for s-risks
You probably won’t successfully build up skills and become capable enough.
I think you point out a good point: Some of the donators are too conservative of donating. I wonder if you have thoughts to this question: https://forum.effectivealtruism.org/posts/n76Hpb8N53JBeeWD4/resolving-paradox-funding-isn-t-bottleneck-vs-80-high
Have you ever considered donating to the field of AI s-risks, like Center for reducing suffering or independent s-risks reserachers?
I think some of Brain Tomasiks essays are quite persuasive: https://briantomasik.com/
Also, I think we could say: Imagine you’re going to be thrown to a volcano for 10 minutes, but you’d get X years of happiness, how many years are you willing to do this exchange? I think most of us even if we wouldn’t want to be thrown to a volcano to exchange 1000 years of happiness, that’s why reducing extreme suffering is important
Some rough ideas- 1.For older people, it’s harder to change their career path to something like AI safety. However, they could still do earn to give(But unfortunately, EtG is less discussed in EA recently)
2.The cost of turning values: It’d be harder for people over 40 to change their value systems they hold for years to completely EA value systems. For me, since I touched EA at 15 so I don’t have this problem.
Thanks very much for your answering, I’m very grateful for it. I think bascially your idea is “The impact of AI risks research is fat-tailed”. But there’s still a question: If there are money left, and 80% of people aren’t funded, why don’t you fund them even though they have little impact? Maybe you’ll say we should save money, but will the AI risks researchers in the future be much more capable than now? In other words, if the funding bar is still that high, is it probable 10 years later, still 80% of people can’t reach the funding bar and do direct work?
It seems many unfunded s-risks researchers are already senior(have 5+ years of experience), which means even if they have 10 years more, they probably wouldn’t get more capable and pass the funding bar. But I’m unceratin and welcome to criticize this idea.
[Question] Resolving Paradox: Funding Isn’t Bottleneck vs. ~80% high Rejection Rates in AI Safety
Could this news really be the evidence of “It’s probably in the future the funding gap would decrease significantly”. Of course in the future 3 years there may be a lot of small donors coming from Anthropic, but what if Anthropic is surpassed by other AI frontier labs in the future?(like: Open AI, Google Deepmind) There may be way fewer donor in these companies. Therefore, the increase of funding may not continue long-term. (Though, I’m very uncertain, welcome to comment below to share your intuitions about this).
It’s interesting and is actually an under discussed but important topic in EA community.
However, I think you could compare direct work vs donating to support AI safety research directly, not donating Givewell(which mainly focus on improving global health) Because for some people, donating to longtermism funds is much more effective than GiveWell.
Hello Jeff,
Thanks a lot for your reply. I’m really grateful for it.
I think you’re right that double majoring in dentistry for an extra 4 years is a risk. For example, if a human-level AGI robot comes in 10 years, dentists might be replaced. However, not studying dentistry could also be a risk if human-level AGI comes slowly — for instance, if AGI isn’t developed in the next 30 years. In that case, dentists probably won’t be replaced for 30 years. But the average CS engineer’s salary may decrease significantly in just 5 years.
You’re right that management is important, but if I’m average at management and I can’t be expert at any useful things, dentistry might be the only path where I can have significant impact (by earning to give).
That’s the reason I’m wondering whether the contribution of an average direct worker may be significantly lower than that of dentists donating $80,000 a year. (However, I’m really uncertain about this. I’d be grateful if you’d like to share your thoughts on it.)
Hello Jeff: (I’m a big fan of your writings).
I’m a college freshman deciding between double majoring in CS and dentistry (8 years total) or majoring only in CS (4 years). Although dentistry isn’t useful for reducing AI risks and isn’t quite interesting to me, the main appeal is adding another earning-to-give route as a dentist.
However, I’m not asking whether I should pursue dentistry. I’d like to isolate one key sub-question here:
If the fat-tailed distribution of impact holds true(as picture below), an average direct worker’s contribution may be negligible compared to the talented (though I’m uncertain). Therefore, if my ability in AI risks direct work turns out average compared to other EA people in future, how would you compare an average direct AI risks worker’s contribution to a dentist who donates an extra $80,000 per year?
Instead of asking which is better, I’d ask: How do you personally evaluate this trade-off?I’ve been thinking for like 300 hours and I feel like hitting diminishing returns from isolated thinking.
Therefore, please don’t aim for a correct or rigorous answer. Replies of 1-minute gut intuitions as short as “My main crux is X” or “I think you may be neglecting Y” would already be extremely helpful. Thank you very much.
Proposal: Train a LLM to be an EA expert
Hello Roman: I’m Jack, I think it’s postively related but not 100% accurate. For example, if an EA organizations is funding constraint, it may decrease its salary and if an EA org is rich it may be more willing to provide high salary. The impact of the former seems not to be less because of lower salary. (Though, working in a funding constraint company may have lower impact compared to donating money to it)
Hello Jonas: I’m Jack, It’s a tough question and I’m also thinking of the answer. However, I want to say your question is really important for EA. For myself, I quit watching NBA(basketball games) everyday and turn to watch NBA paper magazines recently, I found it significantly improved my concentration and productivity. Maybe you can consider check out a topic in EA forum which is called”Productivity” to give you more ideas.
Hello Davey: As I know the closest thing is EA anywhere Slack. You could google it and apply joining it. I resonante with you a lot because I’m also looking for EA poeple to connect and discuss with. However a harsh reality is that it seems most EA people are very busy to reply other’s messages/mails.
Welcome you to contact me by DM me on EA Forum if you have anything related to EA to ask/ discuss. Though I’m definitely not an EA expert, but recently I have time to discuss.
I think this might be actually a good idea, to remind people how fast AGI might arrive
What I mean here is if you can give> $50000 a year, then it could be probable the contribution is better than direct work.
Of course, doing any small altruistic things is valuable and worth admiration(even only giving $1)