Arden Koehler

Thoughts on 80,000 Hours’ re­search that might help with job-search frustrations

• Arden Koehler 17 Apr 2019 21:04 UTC

  23 points

  0 ∶ 0

  in reply to: MichaelPlant’s comment on: Thoughts on 80,000 Hours’ re­search that might help with job-search frustrations

  The tension is between saying ‘our research is useful: it tells (group X) of people what it is best for them to do’ and ‘our research does not offering a definitive ranking of what it is before for people to do (whether people in group X or otherwise)’.

  Though I am saying that 80,000 Hours’ research can’t offer a single, definite ranking of what is best for everyone to do, that doesn’t mean that their research isn’t very useful for people figuring out what it is best for them to do.

  The way I might put it: 80,000 Hours research helps people put together their own list of what is best for them to do, by (1) offering lots of information people need to combine with their own knowledge about themselves to build their list—e.g., what certain jobs are like, how people typically get into a particular job, and so on, (2) offering tools for people to use to figure out the information about themselves that they need—like for assessing personal fit, etc., and (3) offering guidance on how to prioritize options according to the impact that people in the roles can have under various different circumstances. 80,000 hours also does things like seek out specific positions and bring them to people’s attention.

  All this is really useful, I believe, for helping people do the most good they can with their careers, without any of it amounting to creating a big list of what it’s best for everyone in group x (e.g., the EA community) to do.

• Arden Koehler 17 Apr 2019 21:09 UTC

  20 points

  0 ∶ 0

  in reply to: MichaelPlant’s comment on: Thoughts on 80,000 Hours’ re­search that might help with job-search frustrations

  Hey Michael,

  Thanks for commenting. With regard to your first point: I don’t think there is a tension—The idea of a list of the best careers for everyone from top to bottom doesn’t make much conceptual sense. But a list of career paths that it does the most good for a specific set of people to read about and consider does make conceptual sense. I think of 80,000 Hours as more like the latter.

  And as I wrote in a reply to your other comment below, a list like that can be really helpful for people in creating their own personal lists of what the best options are for them.

  (this is basically to agree with cole_haus’s reply)

• Arden Koehler 18 Apr 2019 13:11 UTC

  14 points

  0 ∶ 0

  in reply to: lexande’s comment on: Thoughts on 80,000 Hours’ re­search that might help with job-search frustrations

  If the way you talk about career capital here is indicative of 80k’s current thinking then it sounds like they’ve changed their position AGAIN, mostly reversing their position from 2018

  I didn’t mean for what I said to suggest a departure from 80,000 Hours’ current position on career capital. They still think (and I agree) that it’s better for most people to have a specific plan for impact in mind when they’re building career capital, instead of just building ‘generic’ career capital (generally transferable skills), and that in the best case that means going straight into a career path. But of course sometimes that won’t be possible and people will have to skill up somewhere else.

  how well can people feasibly target narrow career capital 5-10 years out when the skill bottlenecks of the future will surely be different?

  This is a good question, and it’s of course not easy to predict what the most impactful things to do in 5-10 years will be—it seems unlikely, though, that working toward one of 80,000 Hours’ “priority paths” will become not very useful down the line. And in general being sensitive to how skill bottlenecks might change in the future is definitely something that 80,000 Hours is keen on.

  To your second point: I mean, yeah—it’s hard to keep everything up to date, especially as the body of research grows, but it’s obviously bad to have old content up there that is misleading or confusing. Updating and flagging (and maybe removing—I’m not sure) old content is something 80,000 Hours is working on.

  an obvious strategy for dealing with this is to explicitly state (for each page or section of the site) who the intended audience is.

  I’m not sure what exactly the 80,000 Hours team would say about explicitly labeling different pages with notes about the intended audience, but my guess is that they wouldn’t want to do that for a lot of their content because it’s very hard to say exactly who it will be useful for. They do have something about intended audience on their homepage: “Ultimately, we want to help everyone in the world have a big social impact in their career. Right now, we’re focusing on giving advice to talented and ambitious graduates in their twenties and thirties.” I know that’s vague, but it seems like it has to be vague to keep from screening off people who could benefit from the research.

  Maybe they could do a better job of helping people figure out what content is for them and what content isn’t, but it doesn’t seem to me at least like explicitly labels at the top of pages would be the right way to go about it.

• Arden Koehler 19 Apr 2019 3:37 UTC

  10 points

  0 ∶ 0

  in reply to: Vaidehi Agarwalla’s comment on: Thoughts on 80,000 Hours’ re­search that might help with job-search frustrations

  This seems great to me! Thanks for writing this out. As for building flexible career capital (re: the comment below): flexibility is of course good all else equal, and more important the earlier people are in their careers. It’s just that people can face a trade-off at some point between flexibility and usefulness to something specific. I think 80,000 Hours has changed its views on how to weight the considerations in that trade-off, favoring usefulness to something specific more than they used to. But if someone can both work toward something that they think will be really valuable and build flexible career capital at the same time, that seems all the better.

• Arden Koehler 19 Apr 2019 13:17 UTC

  5 points

  0 ∶ 0

  in reply to: lexande’s comment on: Thoughts on 80,000 Hours’ re­search that might help with job-search frustrations

  Hey Lexande-

  Just to address your last point/​question: I don’t think that the right thing to take away from 80,000 Hours changing its mind over the years on some of these points is pessimism about the targeted career capital one builds now being useful in 5-10 years—there are a lot of ways to do good, and these changes reflect 8,000 Hours changing views on what the absolute optimal way of doing good is. That’s obviously a hard thing to figure out, and it obviously changes over time. But most of their advice seems pretty robust to me. Even if it looks like in 5-10 years it would have been absolutely optimal to be doing something somewhat different, having followed 80,000 Hours advice would still likely put you in a position that is pretty close to the best place to be.

  For example, if you are working toward doing governmental AI policy, and in 5-10 years that area is more saturated and so slightly less optimal than they think now it will be, and now it’s better to be working in an independent think tank, or on other technology policy, etc., then (1) what you’re doing is probably still pretty close to optimal, and (2) you might be able to switch over because the direct work you’ve been engaging in has also resulted in useful career capital.

  It’s also important to remember that if in 10 years some 80,000 Hours-recommended career path, such as AI policy, is less neglected than it used to be, that is a good thing, and doesn’t undermine people having worked toward it—it’s less neglected in this case because more people worked toward it.

• Arden Koehler 21 Aug 2019 21:46 UTC

  20 points

  0 ∶ 0

  in reply to: Wei Dai’s comment on: Ask Me Any­thing!

  Hey Wei_Dai, thanks for this feedback! I agree that philosophers can be useful in alignment research by way of working on some of the philosophical questions you list in the linked post. Insofar as you’re talking about working on questions like those within academia, I think of that as covered by the suggestion to work on global priorities research. For instance, I know that working on some of those questions would be welcome at the Global Priorities Institute, and I think FHI would probably also welcome philosophers working on AI questions. But I agree that that isn’t clear from the article, and I’ve added a bit to clarify it.

  But maybe the suggestion is working on those questions outside academia. We mention DeepMind and Open AI as having ethics divisions, but likely only some philosophical questions relevant to AI safely are done in those kinds of centers, and it could be worth listing more non-academic settings in which philosophers might be able to pursue alignment relevant questions. There are, for instance, lots of AI ethics organizations, though most are only focused on short-term issues, and are more concerned with ‘implications’ than with philosophical questions that arise in the course of design. CHAI, AI Impacts, the Leverhulme center, and MIRI also seem to do a bit of philosophy each. The future Schwarzman Center at Oxford may also be a good place for this once it gets going. I’ve edited the relevant sections to reflect this.

  Do you know of any other projects or organizations that might be useful to mention? I also think your list of philosophy questions relevant to AI is useful—thanks for writing it up!-- and would like to link to it in the article.

  As for the comparison with journalism and AI policy, in line with what Will wrote below I was thinking of those as suggestions for people who are trying to get out of philosophy or who will be deciding not to go into it in the first place, i.e., for people who would be good at philosophy but who choose to do something else that takes advantage of their general strengths.

• Arden Koehler 3 Sep 2019 14:45 UTC

  1 point

  0 ∶ 0

  in reply to: Wei Dai’s comment on: Ask Me Any­thing!

  Re: these being alternatives to philosophy, I see what you mean. But I think it’s ok to group together non-academic philosophy and non-philosophy alternatives since it’s a career review of philosophy academia. However, I take the point that I can better connect the two ‘alternatives’ sections in the article and have added a link.

  As for individual grants, I’m hesitant to add that suggestion because I worry that that would encourage some people people who aren’t able to get philosophy roles in academia or in other organizations to go the ‘independent’ route, and I think that will rarely be the right choice.

• Arden Koehler 17 Nov 2019 13:19 UTC

  7 points

  0 ∶ 0

  on: In­sti­tu­tions for Fu­ture Generations

  This is such an exciting project! Really glad you’re doing it.

  I have two questions/​tentative suggestions on the scope/​framing of the project:

  (1) Are you considering any existing institutions? It seems like it could be useful to identify any existing institutions that seem advantageous for future generations—so that we can have a better sense of the value of preserving or expanding them, and in case they could be used as templates for new institutions.

  (2) The evaluation criteria seem good. But should you add something along the lines of “How likely is this institution to gain support in the future from people with nonlongtermist interests in a way that doesn’t undermine its value, such that we wouldn’t need to provide it with as much ongoing support?” (Related to political feasibility but more forward looking—maybe can be just rolled into that criterion.)

Essen­tial facts and figures—COVID-19

• Arden Koehler 24 Apr 2020 17:08 UTC

  1 point

  0 ∶ 0

  on: How hot will it get?

  Great post, and great point about the priors! I have a question about how to use/​interpret these which I’d love help with from you or someone else who understands this better than I do.

  Can I draw implications of your models about emissions scenarios as defined by the IPCC?

  First, can I take the first model to indicate something about how likely various emissions pathways (e.g., RCP 6) are if we take little ‘extra action’? e.g., on the “JH extrapolation” version of business as usual that we’re 95% likely not to reach above the mean RCP 8.5 emissions scenario (6180 Gt), 70% likely not to reach above the mean RCP 6.0 scenario (3885 Gt), etc? (all by 2100)

  Second, can I take your second model to indicate something about how how much warming we’d get if we were to reach those emissions scenarios? So if RCP 6.0 is the 70th percentile outcome of business as usual (on the ‘JH extrapolation’ version), can we then take the 70th percentile of the probability density function for one of the sensitivity assumptions (say, the Webster one) for how hot it will get on that version of business as usual + that sensitivity assumption to get the amount of warming predicted for RCP 6.0 -- i.e., 3C?

• Arden Koehler 29 Apr 2020 14:38 UTC

  1 point

  0 ∶ 0

  in reply to: John G. Halstead’s comment on: How hot will it get?

  Thanks, this is helpful!

Prob­lem ar­eas be­yond 80,000 Hours’ cur­rent pri­ori­ties

• Arden Koehler 22 Jun 2020 19:12 UTC

  1 point

  0 ∶ 0

  in reply to: Pablo’s comment on: Prob­lem ar­eas be­yond 80,000 Hours’ cur­rent pri­ori­ties

  fixed, thanks!

• Arden Koehler 22 Jun 2020 21:05 UTC

  5 points

  0 ∶ 0

  in reply to: MichaelPlant’s comment on: Prob­lem ar­eas be­yond 80,000 Hours’ cur­rent pri­ori­ties

  Hey Michael—there isn’t such a list, though we did consider and decide not to include a number of problems in the process of putting this together. I definately think that “X and Y are on the list so Z, which wasn’t mentioned explicitly, is also likely a good area” would be a bad inference! But there are also probably lots of issues that we didn’t even consider so something not being on the list is probably at best a weak negative signal. [Edit: I shouldn’t have said “at best”—it’s a weak negative signal.]

• Arden Koehler 23 Jun 2020 9:27 UTC

  1 point

  0 ∶ 0

  in reply to: MichaelA’s comment on: Prob­lem ar­eas be­yond 80,000 Hours’ cur­rent pri­ori­ties

  Fixed, thanks!

• Arden Koehler 23 Jun 2020 9:44 UTC

  15 points

  0 ∶ 0

  in reply to: Pablo’s comment on: Prob­lem ar­eas be­yond 80,000 Hours’ cur­rent pri­ori­ties

  Thanks Pablo—I agree we should discuss risks to EA more. It seems like it should be a natural part of ‘building effective altruism’ to me. I wonder why we don’t discuss it more in that area. Maybe people are afraid it will seem self-indulgent?

  I think I’d worry about how to frame it in 80k content because our stuff is very outward-facing and people who aren’t already part of the community might not respond well to it. But that’s less of an issue with forum posts, etc.

  I’d also guess most people’s estimates for EA going away or becoming much less valuable in the next 10 years are lower than yours. Want to expand a bit on why you think it’s as high as you do?

  Thanks for bringing this up and also for the list of places this has been discussed!

• Arden Koehler 23 Jun 2020 9:45 UTC

  3 points

  0 ∶ 0

  in reply to: MichaelA’s comment on: Prob­lem ar­eas be­yond 80,000 Hours’ cur­rent pri­ori­ties

  This is great, thanks!

• Arden Koehler 23 Jun 2020 10:00 UTC

  4 points

  0 ∶ 0

  in reply to: MichaelA’s comment on: Prob­lem ar­eas be­yond 80,000 Hours’ cur­rent pri­ori­ties

  Hey Michael,

  Glad you’ve found it helpful, and thanks for these resource lists! I’m adding them to our inernal list of resources. Anything you’ve read from them you think it’d be particularly good to add to the above blurbs?

• Arden Koehler 24 Jun 2020 17:00 UTC

  3 points

  0 ∶ 0

  in reply to: atlasunshrugged’s comment on: Prob­lem ar­eas be­yond 80,000 Hours’ cur­rent pri­ori­ties

  Hey atlasunshrugged,

  I’m afraid I don’t know the answers to your specific questions. I agree that there are things worse than great power conflict, and perhaps China becoming the dominent world power could be one of those things. FWIW although war between the US and China does seem like one of the more worrying scinarios at the moment, I meant the description problem to be broader than that and include any great power war.