NunoSempere

• NunoSempere 18 Jan 2021 12:00 UTC

  2 points

  on: Fore­cast­ing of Pri­ori­ties: a tool for effec­tive poli­ti­cal par­ti­ci­pa­tion?

  Substantive points

  Starting March 2021, we will test, whether a diversified group of 150-250 people with financial motivations and the basics of forecasting (representing the citizens who would choose the “forecaster” strategy) are actually able to predict the top 5-10 megatrends /​ grand societal challenges (from a long-list of 20-40), that will be prioritized 3-5 years later by experts in a Delphi study.

  Wait, so citizens are incentivized to predict what experts will say? This seems a little bit weak, because experts can be arbitrarily removed from reality. You might think that, no, our experts have a great grasp of reality, but I’d intuitively be skeptical. As in, I don’t really know that many people who have a good grasp of what the most pressing problems of the world are.

  So in effect, if that’s the case, then the key feedback loops of your system are the ones between experts using the Delphi system <> reality, and the loop between experts <> forecasters seems secondary. For example, if I’m asked what Eliezer Yudkowsky will say the world’s top priority is in three years, I pretty much know that he’s going to say “artificial intelligence”, and if you ask me to predict what Greta Thunberg will say, I pretty much know that she’s going to go with “climate change”.

  I think that eventually you’ll need a cleverer system which has more contact with reality. I don’t know how that system would look, though. Perhaps CSET has something to say here. In particular, they have a neat method of taking big picture questions and decomposing them into scenarios and then into smaller, more forecastable questions.

  Anyways, despite this the first round seems like an interesting governance/​forecasting experiment.

  ---

  Also, 150-250 people seems like too little to get great forecasters. If you were optimizing for forecasting accuracy, you might be better off hiring a bunch of superforecasters.

  ---

  Re: Predict-O-Matic problems, see some more here

  ---

  (from a long-list of 20-40) Not allowing forecasters to suggest their own trends (maybe with some very cursory review) seems like an easy mistake to fix.

  Nitpicks:

  The core of the mechanism is a forecasting tool, where each citizen receives a virtual credit (let’s say $200) each year (on their birthday, so that the participation is spread in time). After login, they see a long-list of 20-40 public causes and challenges (such as longevity, corruption, legalization of drugs, mental health, better roads, etc). Citizens can (anonymously) allocate credit anytime within a year, using quadratic voting, to any causes that they consider a priority, and explain why or specify it further in a public comment. They can use different strategies to do that, which I will describe below. Once they allocate the credit, the amount actually goes to solving the cause (i.e. funding research and implementation of the solutions) as funding by the government.

  This may have the problem that once the public identifies a “leader”, either a very good forecaster or a persuasive pundit, they can just copy their forecasts. As a result, this part:

  As a hypothetical result, the government is happy because it effectively harnesses a lot of inputs about what to fund and only pays rewards to the most visionary inputs 3y later.

  seems like an overestimate; you wouldn’t be harnessing that many inputs after all

  ---

  Non-populist politicians are happy because they can tell their voters “your opinion matters” and now it’s believable. The citizens are happy because they feel directly involved in policymaking and get educated in the process. NGOs hoping to improve public discourse are happy because there is a growing structured database of weighted arguments that get checked for accuracy 3 years later. Media are happy because now there is a constantly updating and easily understandable aggregate of what citizens actually think and want.

  This depends on how much of the budget is chosen this way. In the worst case scenario, this gives a veneer of respectability to a process which only lets citizens decide over a very small portion of the budget.

• NunoSempere 18 Jan 2021 10:42 UTC

  2 points

  in reply to: capybaralet’s comment on: Big List of Cause Candidates

  With regards to coral reefs, your post is pretty short. In my experience, it’s more likely that people will pay more attention to it if you flesh it out a little bit more.

• NunoSempere 18 Jan 2021 10:36 UTC

  2 points

  in reply to: capybaralet’s comment on: Big List of Cause Candidates

  Yeah, this makes sense, thanks.

• NunoSempere 16 Jan 2021 10:57 UTC

  2 points

  in reply to: MichaelA’s comment on: A Fun­nel for Cause Candidates

  Makes sense, thanks

• NunoSempere 15 Jan 2021 9:49 UTC

  3 points

  in reply to: MichaelA’s comment on: A Fun­nel for Cause Candidates

  “I have no particular reason to disagree with any of your reasoning about how promising this idea is for the population it intends to benefit, but I just think benefitting that population is [huge number] less important than benefitting [this other population], so I really don’t care.”

  So suppose that the intervention was about cows, and I (the vectors in “1” in the image) gave them some moderate weight X, the length of the red arrow. Then if someone gives them a weight of 0.0001X, their red arrow becomes much smaller (as in 2.), and the total volume enclosed by their cube becomes smaller. I’m thinking that the volume represents promisingness. But they can just apply that division X → 0.0001X to all my ratings, and calculate their new volumes and ratings (which will be different from mine, because cause areas which only affect, say, humans, won’t be affected).

  Or “I have no particular reason to disagree with any of your reasoning about how promising this idea is if we aim to simply maximise expected utility even in Pascalian situations, and accept long chains of reasoning with limited empirical data behind them. But I’m just really firmly convinced that those are bad ways to make decisions and form beliefs, so I don’t really care how well the idea performs from that perspective.”

  In this case, the red arrow would go completely to 0, and that person would just focus on the area of the square in which the blue and green arrows lie, across all cause candidates. Because I am looking at volume and they are looking at areas, our ratings will again differ.

• NunoSempere 14 Jan 2021 12:19 UTC

  2 points

  in reply to: jackmalde’s comment on: A Fun­nel for Cause Candidates

  I have got the impression that there is going to be a single funnelling exercise that aims to directly compare shorttermist vs longtermist areas including on their ‘scale’.

  Yeah, so I (and others) have been exploring different things, but I don’t know what I’ll end up going with. That said, I think that there are gains to be had in optimizing first two stages, not just the third evaluation stage.

• NunoSempere 14 Jan 2021 11:43 UTC

  2 points

  in reply to: NunoSempere’s comment on: A Fun­nel for Cause Candidates

  Nitpick: A change of basis might also be combined with a projection into a subspace. In the example, if one doesn’t care about animals, or about the long term future at all, then instead of the volume of the cuboid they’d just consider the area of one of its faces.

  Another nitpick: The ratio of humans to animals would depend on the specific animals. However, I sort of feel that the high level disagreements of the sort jackmalde is pointing to are probably in the ratio of the value between a happy human life and a happy cow life, not about the ratio of the life of a happy cow to a happy pig, chicken, insect, etc.

• NunoSempere 14 Jan 2021 11:41 UTC

  3 points

  in reply to: jackmalde’s comment on: A Fun­nel for Cause Candidates

  So suppose you have a cause candidate, and some axis like the ones you mention:

  • Degree to which the area arguably runs into the cluelessness problem

  • Degree to which the area affects beings who don’t exist yet /​ Promisingness of the area under the totalist view.

  • ...

  But also some others, like

  • Promisingness given a weight of X humans to Y animals

  • Promisingness given that humans are ~infinitely more valuable than animals

  • Tractability of finding good people to get a project started

  • Units of resources needed

  • Probability that a trusted forecasting system gives to your project still existing in 2 years.

  For simplicity, I’m going to just use three axis, but the below applies to more. Right now, the topmost vectors represent my own perspective on the promisingness of a cause candidate, across three axis, but they could eventually represent some more robust measure (e.g., the aggregate of respected elders, or some other measure you like more). The vectors at the bottom are the perspectives of people who disagree with me across some axis.

  

  For example, suppose that the red vector was “ratio of the value of a human to a standard animal”, or “probability that a project in this cause area will successfully influence the long-term future”.

  Then person number 2 can say “well, no, humans are worth much more than animals”. Or, “well, no, the probability of this project influencing the long-term future is much lower”. And person number 2 can say something like “well, overall I agree with you, but I value animals a little bit more, so my red axis is somewhat higher”, or “well, no, I think that the probability that this project has of influencing the long-term future is much higher”.

  Crucially, they wouldn’t have to do this for every cause candidate. For example, if I value a given animal living a happier life the same as X humans, and someone else values that animal as 0.01X humans, or as 2X humans, they can just apply the transformation to my values.

  Similarly, if someone is generally very pessimistic about the tractability of influencing the long-term future they could transform my probabilities as to that happening. They could divide my probabilities by 10x (or, actually, subtract some amount of probability in bits). Then the tranformation might not be linear, but it would still be doable.

  Then, knowing the various axis, one could combine them to find out the expected impact. For example, one could multiply three axis to get the volume of the box, or add them as vectors and consider the length of the purple vector, or some other transformation which isn’t a toy example.

  So, a difference in perspectives would be transformed into a change of basis

  So:

  Even in that case there’s a difficulty, in that anyone who disagrees with your stance on these foundational questions would then have the right to throw out all of your funneling work and just do their own.

  doesn’t strike me as true. Granted, I haven’t done this yet, and I might never because other avenues might strike me as more interesting, but the possibility exists.

• NunoSempere 14 Jan 2021 9:18 UTC

  2 points

  in reply to: jackmalde’s comment on: A Fun­nel for Cause Candidates

  Makes sense, thanks, changed.

A Fun­nel for Cause Candidates

• NunoSempere 11 Jan 2021 13:26 UTC

  1 point

  in reply to: jackmalde’s comment on: Big List of Cause Candidates

  Acknowledged.

• NunoSempere 11 Jan 2021 12:13 UTC

  1 point

  in reply to: NunoSempere’s comment on: Big List of Cause Candidates

  tl;dr/​Notes:

  I have some models of the world which lead me to think that the idea was unpromising. Some of them clearly have a subjective component. Still, I’m using the same “muscles” as when forecasting, and I trust that those muscles will usually produce sensible conclusions.

  It is possible that in this case I had too negative a view, though not in a way which is clearly wrong (to me). If I was forecasting the question “will a charity be incubated to work on philosophy in schools” (surprise reveal: this is similar to what I was doing all along), I imagine I’d give it a very low probability, but that my team mates would give it a slightly higher probability. After discussion, we’d both probably move towards the center, and thus be more accurate.

  Note that if we model my subjective promisingness = true promisingness + error term, if we pick the candidate idea at the very bottom of my list (in this case, philosophy in schools, the idea under discussion and one of the four ideas to which I assigned a “very unpromising” rating), we’d expect it to both be unpromising (per your own view) and have a large error term (I clearly don’t view philosophy very favorably)

• NunoSempere 11 Jan 2021 11:56 UTC

  1 point

  in reply to: jackmalde’s comment on: Big List of Cause Candidates

  Let me try to translate my thoughts to something which might be more legible /​ written in a more formal tone.

  • From my experience observing this in Spain, the philosophy curriculum taught in schools is a political compromise, in which religion plays an important role. Further, if utilitarism is even taught (it wasn’t in my high school philosophy class), it can be taught badly by proponents of some other competing theory. I expect this to happen, because most people (and by expectation most teachers) aren’t utilitarian.

  • Philosophy doesn’t have high epistemic standards, as evidenced by the fact that it can’t come to a conclusion about “who is right”. Some salient examples of philosophers who continue to be taught and given significant attention despite having few redeeming qualities are Plotinous, Anaximenes, or Hegel. Although it can be argued that they do have redeeming qualities (Anaximenes was an early proponent of proto-scientific thinking, and Hegel has some interesting insights about history, and has shaped further thought), paying too much attention to these philosophers would be the equivalent of coming to deeply understand phologiston or aether theory when studying physics. I understand that grading the healthiness of a field can be counterintuitive or weird, but to the extent that a field can be sick, I think that philosophy ranks near the bottom (in contrast, development economics of the sort where you do an RCT to find out if you’re right would be near the top)

  • Relatedly, when teaching philosophy too much attention is usually given to the history of philosophy. I agree that an ideal philosophy course which promoted “critical thinking” would be beneficial, but I don’t think that it would be feasible to implement it because: a) it would have to be the result of tricky political compromise and have to be very careful around critizicing whomever is in power, and b) I don’t think that there are enough good teachers who could pull it off.

  • Note that I’m not saying that philosophy can’t produce success stories, or great philosophers, like Parfit, David Pearce, Peter Singer, arguably Bostrom, etc (though note that all examples except Singer are pretty mathematical). I’m saying that most of the time, the average philosophy class is pretty mediocre

  • On this note, I believe that my own (negative) experience with philosophy in schools is more representative than yours. Google brings up that you went to Cambridge and UCL, so I posit that you (and many other EAs who have gone to top universities) have an inflated sense of how good teachers are (because you have been exposed to smart and at least somewhat capable teachers, who had the pleasure of teaching top students). In contrast, I have been exposed to average teachers who sometimes tried to do the best they could, and who often didn’t really have great teaching skills.

• NunoSempere 11 Jan 2021 10:35 UTC

  2 points

  in reply to: BrianTan’s comment on: 2020: Fore­cast­ing in Review

  Thanks!

• NunoSempere 11 Jan 2021 10:34 UTC

  1 point

  in reply to: jackmalde’s comment on: Big List of Cause Candidates

  I can imagine reconsidering, but I don’t in principle have anything against using my S1. Because:

  • It is fast, and I am rating 100+ causes

  • From past experience with forecasting, I basically trust it.

  • It does in fact have useful information. See here for some discussion I basically agree with.

• NunoSempere 10 Jan 2021 22:37 UTC

  3 points

  in reply to: jackmalde’s comment on: Big List of Cause Candidates

  OK this seems fairly personal and anecdotal

  Yeah, this is fair. Ideally I’d ask a bunch of people what their subjective promisingness was, and then aggregate over that. I’d have to somehow adjust for the fact that people from EA backgrounds might have gone to excellent universities and schools, and thus their estimate of teacher quality might be much, much higher than average, though.

• NunoSempere 10 Jan 2021 22:17 UTC

  2 points

  in reply to: vaidehi_agarwalla’s comment on: Why EA meta, and the top 3 char­ity ideas in the space

  Well, it’s not like we’re each adversarially trying to maximize our own counterfactual impact, rather than impact as community :P

• NunoSempere 10 Jan 2021 22:14 UTC

  7 points

  on: Why EA meta, and the top 3 char­ity ideas in the space

  Here is a past list of EA charity ideas/​cause candidates I collected from a previous project, organized by whether they are meta or partially meta (i.e., “a step removed from direct impact”), and then by my subjective promisingness. You can see more information by clicking on the “expand” button near each cause area. They come from a recent project some other people have linked to.

  Some of the ones which I think you might think are competitive with your top three ideas are:

  • Increasing development aid

  • Getting Money Out Of Politics And Into Charity

  • Non-Western EA

  • For-profit companies serving emerging markets

  Additionally, some of the ones which I think might be competitive with your top three ideas (but about which you might disagree) are:

  • Epistemic Progress

  • Patient Philanthropy

  • Institutions for Future Generations

  • Counter-Cyclical Donation Timing

  • Effective Informational Lobbying

  You can see more in the link above. I also have “Discovering Previously Unknown Existential Risks”, which is pretty related to your “Exploratory altruism” cause, and “Effective Animal Advocacy Movement Building” (in which cause CE has incubated Animal Advocacy careers).

• NunoSempere 10 Jan 2021 20:55 UTC

  2 points

  in reply to: Prabhat Soni’s comment on: Big List of Cause Candidates

  Ok, cheers, will add.

• NunoSempere 10 Jan 2021 20:54 UTC

  2 points

  on: Big List of Cause Candidates

  To do:

  • Add https://​​forum.effectivealtruism.org/​​posts/​​MjWHupe8d8mMGJqKP/​​ea-and-the-possible-decline-of-the-us-very-rough-thoughts as a cause candidate.

  • Add risks from whole Brain emulation.

  • Add preventing/​avoiding stable longterm totalitarianism.

  • Add atomically precise manufacturing /​ molecular nanotechnology

  • Add https://​​forum.effectivealtruism.org/​​posts/​​FAA22RbfgC68fRnRs/​​if-you-mostly-believe-in-worms-what-should-you-think-about