What posts do you want someone to write?

post by Aaron Gertler (aarongertler) · 2020-03-24T06:41:35.531Z · score: 52 (18 votes) · EA · GW · No comments

This is a question post.

Contents

  Answers
    28 bmg
    23 Aaron Gertler
    18 EdoArad
    18 Aaron Gertler
    16 technicalities
    11 EdoArad
    11 Aaron Gertler
    10 Ben_West
    10 Ben_West
    8 RomeoStevens
    7 Milan_Griffes
    6 Aaron Gertler
    5 Derek
    5 Mati_Roy
    5 technicalities
    4 Aaron Gertler
    3 evelynciara
    3 sky
    2 evelynciara
    1 Alexis Carlier
    1 MichaelA
None
No comments

I really enjoyed "What posts are you planning on writing? [EA · GW]"

This is the lazy version, for people who want a post to exist but want someone else to write it. Given that we're all stuck inside anyway, I'm hoping we can use this opportunity to get a lot of writing done (like Isaac Newton!)

So: What are some unwritten posts that you think ought to exist?

If you want something to exist badly enough that you'd pay for it, consider sharing that information! There's nothing wrong with offering money; some really good work [EA · GW] has come from small-scale contracting of this kind.

Answers

answer by Ben Garfinkel (bmg) · 2020-03-26T21:34:48.791Z · score: 28 (13 votes) · EA(p) · GW(p)

I'd be really interested in reading an updated post that makes the case for there being an especially high (e.g. >10%) probability that AI alignment problems will lead to existentially bad outcomes.

There still isn't a lot of writing explaining case for existential misalignment risk. And a significant fraction of what's been produced since Superintelligence is either: (a) roughly summarizing arguments in Superintelligence, (b) pretty cursory, or (c) written by people who are relative optimists and are in large part trying to explain their relative optimism.

Since I have the (possibly mistaken) impression that a decent number of people in the EA community are quite pessimistic regarding existential misalignment risk, on the basis of reasoning that goes significantly beyond what's in Superintelligence, I'd really like to understand this position a lot better and be in a position to evaluate the arguments for it.

(My ideal version of this post would probably assume some degree of familiarity with contemporary machine learning, and contemporary safety/robustness issues, but no previous familiarity with arguments that AI poses an existential risk.)

comment by mike_mclaren · 2020-04-04T12:12:14.185Z · score: 3 (2 votes) · EA(p) · GW(p)

I'd be really interested in reading an updated post that makes the case for there being an especially high (e.g. >10%) probability that AI alignment problems will lead to existentially bad outcomes.

My understanding is that Toby Ord does just this in his new book The Precipice (his new AI x-risk estimate is also discussed in his recent 80K podcast interview about the book), though it would still be good to have others weigh in.

comment by Ben Garfinkel (bmg) · 2020-04-04T23:55:55.191Z · score: 7 (4 votes) · EA(p) · GW(p)

I think that chapter in the Precipice is really good, but it's not exactly the sort of thing I have in mind.

Although Toby's less optimistic than I am, he's still only arguing for a 10% probability of existentially bad outcomes from misalignment.* The argument in the chapter is also, by necessity, relatively cursory. It's aiming to introduce the field of artificial intelligence and the concept of AGI to readers who might be unfamiliar with it, explain what misalignment risk is, make the idea vivid to readers, clarify misconceptions, describe the state of expert opinion, and add in various other nuances all within the span of about fifteen pages. I think that it succeeds very well in what it's aiming to do, but I would say that it's aiming for something fairly different.

*Technically, if I remember correctly, it's a 10% probability within the next century. So the implied overall probability is at least somewhat higher.

comment by mike_mclaren · 2020-04-05T11:58:47.800Z · score: 2 (2 votes) · EA(p) · GW(p)

I see, thanks for the explanation!

answer by Aaron Gertler · 2020-03-24T07:03:16.309Z · score: 23 (16 votes) · EA(p) · GW(p)

More journalistic articles about EA projects.

I don't necessarily mean "written by journalists", though there's been a lot of good journalistic coverage of EA.

I mean "in the style of long-form journalism": Telling an interesting story about the work of a person/organization, while mixing in the origin story, interesting details about the people involved, photos, etc.

Examples of projects I think could get the journalistic treatment:

comment by Aidan O'Gara · 2020-03-24T07:42:34.185Z · score: 1 (1 votes) · EA(p) · GW(p)

That’s a super cool idea.

  • What writing currently exists like this? Vox’s Future Perfect, maybe a few one-off articles in other major publications?
  • Where’s best to publish this? Feels like a lot of work for a blogpost, but I doubt the NYT is looking for unsolicited submissions - are there publishing platforms that would be interested in this?
comment by Aaron Gertler (aarongertler) · 2020-03-31T11:17:10.567Z · score: 2 (1 votes) · EA(p) · GW(p)

What writing currently exists like this?

Future Perfect and a few one-off articles, mostly. Tom Chivers is a journalist with strong EA leanings who routinely writes from that perspective.

Where's best to publish this?

I wasn't thinking that these stories would have to be published by a large media outlet; I just want them to exist somewhere so that I can share them with people who are new to the movement. 

Getting published on a wider platform could be great for certain orgs (e.g. Wave is just a business, I imagine they wouldn't mind the attention), but bad for others (CSET generally keeps its work fairly private). I'd hope that anyone writing one of these hypothetical stories would check the org's publicity preferences before submitting a story anywhere!

comment by MichaelA · 2020-03-24T08:02:09.818Z · score: 1 (1 votes) · EA(p) · GW(p)

I read in as-yet-unpublished post that the best approach for getting published in a major outlet without being on their staff is not to just write something and then send it to various publications, but rather to pick an outlet and optimise the piece (or versions of it) for that outlet's style, topic choices, readership, etc. (I'm not sure what the evidence base for that claim was, and have 0 relevant knowledge of my own.)

If that is a good approach, one could still potentially pick a few outlets and write somewhat different versions for each, rather than putting all their eggs in one basket. Or write one optimised version at a time, and not invest additional effort until that one is rejected. And one version could also be posted to the EA Forum and/or Medium and/or similar places, in the meantime. (Unless that would reduce odds of publication by a major outlet?)

comment by Aidan O'Gara · 2020-03-25T05:54:48.717Z · score: 1 (1 votes) · EA(p) · GW(p)

Makes a lot of sense, I'm sure Vox and the New York Times are interested in very different kinds of submissions, writing with a particular style in mind probably dramatically increases the odds of publication.

I still wonder what the success rate here is - closer to 1% or to 10%? If the latter, I could see this being pretty impactful and possibly scalable.

answer by EdoArad · 2020-03-24T16:12:22.821Z · score: 18 (11 votes) · EA(p) · GW(p)

Governance innovation as a cause area

Many people are working on new governance mechanisms from an altruistic perspective. There are many sub-categories such as Charter cities [EA · GW], space governance [EA · GW], decentralized governance,  RadicalXChange agenda..

I'm uncertain as to the marginal value in such projects, and I'd like to see a broad analysis that can serve as a good prior and analysis framework for specific projects.

comment by Alexis Carlier · 2020-04-03T17:35:18.268Z · score: 4 (3 votes) · EA(p) · GW(p)

I doubt that there is any one answer re the marginal value of such projects, because the value depends on what is being governed. For instance, I think a successful implementation of regulatory markets for AI safety would be very valuable, but regulatory markets for corporate law wouldn't be; yet the same basic framework is being implemented.

For this reason, I'd be more interested in analysis of governance innovation for a particular cause area.

comment by Halffull · 2020-03-25T15:18:24.150Z · score: 4 (1 votes) · EA(p) · GW(p)

Here's an analysis by 80k. https://80000hours.org/problem-profiles/improving-institutional-decision-making/

comment by EdoArad (edoarad) · 2020-03-25T19:44:11.587Z · score: 1 (1 votes) · EA(p) · GW(p)

This is not quite what I was going for, even though it is relevant. This problem profile focuses on existing institutions and on methods for collective decision making. I was thinking more in the spirit of market design, where the goal is to generate new institutions with new structures and rules so that people are selfishly incentivised to act in a way which maximizes welfare (or something else).

comment by Halffull · 2020-03-25T20:58:04.145Z · score: 3 (2 votes) · EA(p) · GW(p)

I think the framing is weird because of EAs allergy to systemic change, but I think on practice all of the points in that cause profile apply to broader change.

comment by EdoArad (edoarad) · 2020-03-26T06:17:14.114Z · score: 1 (1 votes) · EA(p) · GW(p)

No, the analysis does not seem to contain what I was going for. 

Curious about what you think is weird in the framing?

comment by Halffull · 2020-03-26T15:51:04.051Z · score: 1 (1 votes) · EA(p) · GW(p)
Curious about what you think is weird in the framing?

The problem framing is basically spot on, talking about how our institution drive our lives. Like I said, basically all the points get it right and apply to broader systemic change like RadX, DAOs, etc.

Then, even though the problem is framed perfectly, the solution section almost universally talks about narrow interventions related to individual decision making like improving calibration.

answer by Aaron Gertler · 2020-03-24T06:56:41.008Z · score: 18 (10 votes) · EA(p) · GW(p)

I want a post on how to be a good donor.

Context: I work with a small foundation that asks a lot of questions when we investigate charities. We sometimes worry that we're annoying the charities we work with without providing much value for them or for ourselves, especially since we don't make grants on the same scale as larger foundations. Even when they tell us our questions are helpful/reasonable, they obviously have a strong incentive to make us feel happy and valued. 

Ideal version of this post: Someone goes to a lot of EA orgs, asks them questions related to the above dilemma, and reports the results. 

Other general questions about "what donors should know" would also be neat: How should someone with no special preferences time their donations? How much more valuable is unrestricted than restricted funding? And so on.

comment by Sanjay · 2020-03-24T09:58:46.262Z · score: 6 (4 votes) · EA(p) · GW(p)

This commented was pointed out to me by someone who thought I may be extremely well qualified to answer this.

I have been on the trustee boards of about half a dozen charities and performed short term consulting stints (about a month at a time) for another 10(ish) charities globally, and have seen how each of those engages with its donors.

I'm also know people people/organisations surveying charities about questions like how much more valuable is unrestricted than restricted funding.

I would be happy to put something together on this topic, however I'm snowed under with other things for the time being, but could add it to the list and tackle it later?

comment by MichaelA · 2020-03-24T08:48:00.920Z · score: 3 (2 votes) · EA(p) · GW(p)

I also think this'd be useful.

Though I wonder why you suggest that someone should ask these questions to a lot of EA orgs in particular? Did you also mean orgs that aren't explicitly "EA orgs" but that many EAs see as high-value donation opportunities? And is it possible it'd also be valuable to ask non-EA foundations about their practices and thoughts on this matter, at least as an interesting quite different example?

comment by Aaron Gertler (aarongertler) · 2020-03-24T08:50:34.264Z · score: 3 (2 votes) · EA(p) · GW(p)

If people want to ask other charities, that also seems fine! I suppose I was assuming that EA charities probably do more engagement with small donors (in the sense of "answering lots of questions about their work") than most other charities, and that they might be easier to contact for someone who reads the Forum and sees my post. But I'd guess there would be more value in having a wider sample of organizations.

comment by MichaelA · 2020-03-24T08:57:49.615Z · score: 1 (1 votes) · EA(p) · GW(p)

That all seems to make sense.

answer by technicalities · 2020-03-24T08:35:03.042Z · score: 16 (9 votes) · EA(p) · GW(p)

A case study of the Scientific Revolution in Britain as intervention by a small group. This bears on one of the most surprising facts: the huge distance, 1.5 centuries, between the scientific and industrial revs. Could also shed light on the old marginal vs systemic argument: a synthesis is "do politics - to promote nonpolitical processes!"

https://forum.effectivealtruism.org/posts/RfKPzmtAwzSw49X9S/open-thread-46?commentId=rWn7HTvZaNHCedXNi [EA(p) · GW(p)]

answer by EdoArad · 2020-03-24T15:59:44.109Z · score: 11 (7 votes) · EA(p) · GW(p)

An analysis of how knowledge is constructed in the EA community, and how much weight we should assign to ideas "supported by EA". 

The recent question on reviews by non-EA researchers  [EA · GW] is an example of that. There might be great opportunities to improve EA intellectual progress.

comment by Ikaxas · 2020-03-24T20:18:15.185Z · score: 1 (1 votes) · EA(p) · GW(p)

Ooh, I would also very much like to see this post

answer by Aaron Gertler · 2020-03-24T06:51:42.128Z · score: 11 (8 votes) · EA(p) · GW(p)

An AMA from someone who works at a really big foundation that leans EA but isn't quite "EA-aligned" in the same way as Open Philanthropy (e.g. Gates, Rockefeller, Chan/Zuckerberg, Skoll).

I'm interested to hear how those organizations compare different causes, distribute resources between areas/divisions, evaluate the impact of their grantmaking, etc.

comment by Aidan O'Gara · 2020-03-24T07:47:09.930Z · score: 7 (5 votes) · EA(p) · GW(p)

Similarly, an AMA from someone working at an EA org who otherwise isn’t personally very engaged with EA. Maybe they really disagree with EA, or more likely, they’re new to EA ideas and haven’t identified with EA in the past.

They’ll be deeply engaged on the substantive issues but will bring different identities and biases, maybe offering important new perspectives.

answer by Ben_West · 2020-03-24T16:54:08.791Z · score: 10 (7 votes) · EA(p) · GW(p)

Defining "management constraints" better.

Anecdotally, many EA organizations seem to think that they are somehow constrained by management capacity. My experience is that this term is used in different ways (for example, some places use it to mean that they need senior researchers who can mentor junior researchers; others use it to mean that they need people who can do HR really well).

It would be cool for someone to interview different organizations and get a better sense of what is actually needed here

answer by Ben_West · 2020-03-24T15:31:39.558Z · score: 10 (6 votes) · EA(p) · GW(p)

More accessible summaries of technical work. Some things I would like summarized:

1. Existential risk and economic growth [EA · GW]
2. Utilitarianism with and without expected utility

(You can see my own attempt to summarize something similar to #2 here [EA · GW] , as one example.)

comment by Alex HT · 2020-03-31T14:26:10.092Z · score: 6 (4 votes) · EA(p) · GW(p)

I'm thinking of doing (1). Is there a particular way you think this should look?

How technical do you think the summary should be? The thing that would be easiest for me to write would require some maths understanding (eg. basic calculus and limits) but no economics understanding. Eg. about as technical as your summary but more maths and less philosophy.

Also, do you have thoughts on length? Eg. do you think a five page summary is substantially more accessible than the paper, or would the summary have to be much shorter than that?

(I'm also interested in what others would find useful)

comment by Ben_West · 2020-04-01T20:42:10.345Z · score: 2 (1 votes) · EA(p) · GW(p)

Awesome!

I personally would suggest a format of:
1. One paragraph summary that any educated layperson easily can understand
2. One page summary that a layperson with college-level math skills can understand
3. 2-5 pages of detail that someone with college-level math and Econ 101 skills can understand

This is just a suggestion though, I don't have a lot of confidence that it's correct.

comment by Alex HT · 2020-04-30T15:06:21.376Z · score: 10 (7 votes) · EA(p) · GW(p)

Now done here [EA · GW]. It's a ~10 page summary that someone with college-level math can understand (though I think you could read it, skip the math, and get the general idea).

comment by Ben_West · 2020-04-30T17:34:53.740Z · score: 2 (1 votes) · EA(p) · GW(p)

You rock, thanks so much!

comment by MichaelStJules · 2020-07-25T06:59:14.004Z · score: 6 (3 votes) · EA(p) · GW(p)

On 2, see this post [EA · GW] (a link post for this).

I also left some comments on the EA Forum post pulling out the first two theorems and the definitions to state them in way that's hopefully a bit more accessible, skipping some unnecessary jargon and introducing notation only just before it's used, rather than at the start so you have to jump back. They're still pretty technical, though. Upon reflection, it probably took me more time to write the comments than it'll save people to read my comments instead of reading the parts of the paper where they're found. :/

There are also several other theorems in that paper.

comment by Ben_West · 2020-07-29T21:55:20.041Z · score: 2 (1 votes) · EA(p) · GW(p)

Thanks! Yeah I should have posted that both of these have now been published, so if anyone else reading this has a request for posts that they haven't stated publicly, consider doing so!

answer by RomeoStevens · 2020-03-28T02:26:39.194Z · score: 8 (2 votes) · EA(p) · GW(p)

"Type errors in the middle of arguments explain many philosophical gotchas: 10 examples"

"CNS imaging: a review and research agenda" (high decision relevance for moral uncertainty about suffering in humans and non humans)

"Matching problems: a literature review"

"Entropy for intentional content: a formal model" (AI related)

"Graph traversal using negative and positive information, proof of divergent outcomes" (neuroscience relevant potentially)

"One weird trick that made my note taking 10x more useful"

comment by EdoArad (edoarad) · 2020-03-28T08:14:33.668Z · score: 3 (2 votes) · EA(p) · GW(p)

Do you mind expanding a bit on CNS Imaging, Entropy for Intentional content, and Graph Traversal?

answer by Milan_Griffes · 2020-03-28T18:01:31.579Z · score: 7 (7 votes) · EA(p) · GW(p)

"American UBI: for and against"

"A brief history of Rosicrucianism & the Invisible College"

"Were almost all the signers of the Declaration of Independence high-degree Freemasons?"

"Have malaria case rates gone down in areas where AMF did big bednet distributions?"

"What is the relationship between economic development and mental health? Is there a margin at which further development decreases mental health?"

"Literature review: Dunbar's number"

"Why is Rwanda outperforming other African nations?"

"The longtermist case for animal welfare"

"Philosopher-Kings: why wise governance is important for the longterm future"

"Case studies: when has democracy outperform technocracy? (and vice versa)"

"Examining the tradeoff between coordination and coercion"

"Spiritual practice as an EA cause area"

"Tools for thought as an EA cause area"

"Is strong, ubiquitous encryption a net positive?"

"How important are coral reefs to ocean health? How can they be protected?"

"What role does the Amazon rainforest play in regulating the North American biosphere?"

"What can the US do to protect the Amazon from Bolsonaro?"

"Can the Singaporean governance model scale?"

"Is EA complacent?"

"Flow-through effects of widespread addiction"

comment by Derek · 2020-04-03T16:30:39.643Z · score: 3 (2 votes) · EA(p) · GW(p)

"The longtermist case for animal welfare"

Have you seen this? https://forum.effectivealtruism.org/posts/W5AGTHm4pTd6TeEP3/should-longtermists-mostly-think-about-animals [EA · GW]

comment by Milan_Griffes · 2020-04-03T18:58:34.363Z · score: 3 (2 votes) · EA(p) · GW(p)

I hadn't, thanks!

answer by Aaron Gertler · 2020-03-31T11:10:57.544Z · score: 6 (4 votes) · EA(p) · GW(p)

A detailed study of hyper-competent ops people. 

What makes these people so competent? What tools and processes do they use to manage information and set priorities? What does the flow of their workday look like; mostly flitting around between tasks, or mostly focused blocks of time? (And so on.)

answer by Derek · 2020-04-03T16:33:26.962Z · score: 5 (3 votes) · EA(p) · GW(p)

Should Covid-19 be a priority for EAs?

A scale-neglectedness-tractability assessment, or even a full cost-effectiveness analysis, of Covid as a cause area (compared to other EA causes) could be useful. I'm starting to look into this now – please let me know if it's already been done.

answer by Mati_Roy · 2020-03-25T10:03:21.719Z · score: 5 (3 votes) · EA(p) · GW(p)

Negative income taxes > UBI ?

A short mathematical demonstration of how negative income taxes compares to UBI in terms of economics 101.

Here's a thread in an EA group about the topics

answer by technicalities · 2020-03-24T08:40:03.140Z · score: 5 (4 votes) · EA(p) · GW(p)

Collating predictions made by particularly big pundits and getting calibration curves for them. Bill Gates is getting a lot of attention now for warning of pandemic in 2015; what is his average though? (This is a bad example though, since I expect his advisors to be world-class and to totally suppress his variance.)

If this could be hosted somewhere with a lot of traffic, it could reinforce good epistemics.

answer by Aaron Gertler · 2020-03-24T06:48:57.655Z · score: 4 (3 votes) · EA(p) · GW(p)

A post about when we should and should not use "lives saved" language in describing EA work.

I find that telling people they can save a life for $5000 often leads to a lot of confusion: Whose life is being saved? What if they die of something else a few months later?Explaining QALYs isn't too hard if you have a couple of minutes, but you often have a lot less time than that.

Is there some shorthand we can use for "giving 50 healthy years, in expectation, across a population" that makes it sound anywhere nearly as good as simply "saving a life"? How important is it to be accurate in this dimension, vs. simply allowing people to conflate QALY/VSL with "saving a specific person"?

answer by evelynciara · 2020-08-06T06:03:06.467Z · score: 3 (2 votes) · EA(p) · GW(p)

I'd appreciate a forum post or workshop about how to interpret empirical evidence. Jennifer Doleac gives a lot of good pointers in the recent 80,000 Hours podcast, but I think the EA and public policy communities would benefit from a more thorough treatment.

answer by sky · 2020-03-29T13:18:27.270Z · score: 3 (2 votes) · EA(p) · GW(p)

Posts on how people came to their values, how much individuals find themselves optimizing for certain values, and how EA analysis is/isn't relevant. Bonus for points for resources for talking about this with other people.

I'd like to have more "Intro to EA" convos that start with, "When I'm prioritizing values like [X, Y, Z], I've found EA really helpful. It'd be less relevant if I valued [ABC ] instead, and it seems less relevant in those times when I prioritize other things. What do you value? How/When do you want to prioritize that? How would you explore that?"

I think personal stories here would be illustrative.

comment by technicalities · 2020-03-29T14:08:06.017Z · score: 2 (2 votes) · EA(p) · GW(p)

A nice example of the second part, value dependence, is Ozy Brennan's series reviewing GiveWell charities.

Why might you donate to GiveDirectly?
You need a lot of warmfuzzies in order to motivate yourself to donate.
You think encouraging cash benchmarking is really important, and giving GiveDirectly more money will help that.
You want to encourage charities to do more RCTs on their programs by rewarding the charity that does that most enthusiastically.
You care about increasing people’s happiness and don’t care about saving the lives of small children, and prefer a certainty of a somewhat good outcome to a small chance of a very good outcome.
You believe, in principle, that we should let people make their own decisions about their lives.
You want an intervention that definitely has at least a small positive effect.
You have just looked at GDLive and are no longer responsible for your actions.
answer by evelynciara · 2020-03-26T15:05:41.706Z · score: 2 (2 votes) · EA(p) · GW(p)

I care about a lot of different U.S. policy issues and would like to get a sense of their neglectedness and tractability. So I'd love it if someone could do a survey to find out how many people in the U.S. work full time on various issues and how hard it is to get bills passed on them.

answer by Alexis Carlier · 2020-04-03T18:15:10.292Z · score: 1 (3 votes) · EA(p) · GW(p)

When to use quantitative vs qualitative research

MacAskill mentions some considerations here [EA(p) · GW(p)], but the dividing line still feels fuzzy. Sample size is one consideration, but I suspect there are many others, such as the goal of the research (e.g. arguing for the possibility vs the plausibility of some phenomena).

This is relevant to many EA questions, especially those relating to longtermism or disruptive technologies. For instance, this post [LW · GW] uses qualitative methods (in depth case studies) to argue that "an AI which is generally more intelligent than us could take over the world, even if it isn't superintelligent." I'm unsure whether three case studies actually constitutes much evidence; in a comment [LW(p) · GW(p)], the author suggests that a higher-n study ("quantitative") would be helpful.

Without a framework for thinking about this, I'm often unsure what I should be learning from qualitative studies, and I don't know when it makes sense to conduct them. (This seems related to the debate between cleometricians and counterfactual narrative historians; some discussion here, page 18)

comment by Khorton · 2020-04-03T20:36:39.459Z · score: 3 (2 votes) · EA(p) · GW(p)

I don't think this can be taught in one post, because you have to be able to actually use the research methods before you can decide which one to use.

answer by MichaelA · 2020-03-27T06:10:55.778Z · score: 1 (1 votes) · EA(p) · GW(p)

Posts investigating/discussing any of the questions listed here [EA · GW]. These are questions which would be "valuable for someone to research, or at least theorise about, that the current pandemic in some way 'opens up' or will provide new evidence about, and that could inform EAs’ future efforts and priorities".

If anyone has thought of such questions, please add them as answers to to that post.

An example of such a question which I added: "What lessons can be drawn from [events related to COVID-19] for how much to trust governments, mainstream experts, news sources, EAs, rationalists, mathematical modelling by people without domain-specific expertise, etc.? What lessons can be drawn for debates about inside vs outside views, epistemic modesty [EA · GW], etc.?"

No comments

Comments sorted by top scores.