NASA will re-direct an asteroid tonight as a test for planetary defence (link-post) 2022-09-26T04:58:47.354Z
Cause exploration prize: organophosphate pesticides and other neurotoxicants 2022-08-09T10:24:40.697Z
Simulated annealing, or, the importance of informal EA socialising 2022-07-28T05:08:59.547Z
A thinking tool for whether X causes Y 2022-04-12T04:48:34.022Z
Tolstoy's 'Master and Man': a moving short story with an EA message 2022-02-19T03:55:21.660Z
[Creative Writing Contest] Noumenon 2021-10-27T10:50:11.688Z
Does sequence obfuscation present a bio-threat? Probably not (yet) 2021-09-03T09:33:33.533Z


Comment by BenStewart on [deleted post] 2022-09-21T23:02:32.987Z

Hi there! I've just finished my medical degree after an undergraduate in philosophy/int. relations/neuroscience, so really interested in bioethics. I'm going into charity entrepreneurship next year so will be working in the global health/health security space over the coming years

Comment by Ben Stewart (BenStewart) on Cause exploration: Tobacco harm reduction · 2022-09-19T00:19:12.735Z · EA · GW

I think you would need to engage much more fully with gateway effect arguments. Those arguments are strongly proposed by some domain experts, and bear directly on the intervention (Becky Freeman's work was my introduction - but here's a systematic review and meta-analysis). The rates of those who go from no-smoking-->e-cigarettes-->smoking are significant. Even if they are smaller than the rates of those who go from smoking-->e-cigarettes-->no-smoking, they would reduce the overall benefits of THR. And they are especially concerning when considering effects on adolescent cohorts (here's another systematic review and meta-analysis).

The response you cite is from an org with ties to tobacco companies,  by an author who is substantially funded by tobacco companies. I think a higher degree of scepticism is warranted. Also,  consider the strategic implications: would Phillip Morris be pushing so hard in the area if they thought e-cigarettes led to dramatic rates of smoking cessation?

Comment by Ben Stewart (BenStewart) on Cause exploration prize: organophosphate pesticides and other neurotoxicants · 2022-09-18T03:03:33.253Z · EA · GW

I broadly agree - I think for any realistic analysis of this area, the question isn't whether regulation completely solves the problem, but what relative reduction is achievable. I tried to model this in the BOTEC by being fairly conservative about the decrease in exposure:

However, these inputs are also accounting for the probability of success of passing regulation at all - so it's reasonable to think these estimates should be even more conservative. 

This is action-guiding too - selecting which countries to work in would need to consider the domestic regulatory environment  and what the supply chain looks like in that context.

And I think your point is supported by my impression of the experiences of the Lead Exposure Elimination Project. Their impact in Malawi, their first target country, wasn't through passing a law, but in getting pre-existing standards enforced and monitored. If any org was pursue organophosphates as a cause area in the future, I think assisting monitoring and compliance would be an important component. 

Comment by Ben Stewart (BenStewart) on Cause exploration prize: organophosphate pesticides and other neurotoxicants · 2022-09-18T02:49:03.581Z · EA · GW

Great point

Comment by Ben Stewart (BenStewart) on Cause exploration prize: organophosphate pesticides and other neurotoxicants · 2022-09-16T03:43:53.824Z · EA · GW

Thanks alene! These are interesting points and are important to think about. I agree there's a decent chance that insects are sentient (maybe 0.1 to 20% range with extreme uncertainty, and variance depending on which specific species we are talking about). The best source I know on this is Rethink Priorities' report, with exhaustive results here, and the bottom line opinion of each author summarised here. The number of insects and potential for suffering is gravely concerning.

However, I don't think that justifies a complete opposition to pesticide use. Pesticides are extremely useful and valuable for feeding more humans and reducing the amount of land needed for crops. Washuck et al. (2022) analysed studies of pesticide use in the U.S. and Canada 2015-2019. They calculated that pesticides preserved millions of hectares of habitats and fed millions more humans (table below).

But there's a deep discussion about ecology, climate change, food systems, agriculture, wild animal welfare and much else that besides that complicates the above; I really don't know. 

As well as  the large benefits of pesticides, the other reason I would tentatively oppose a blanket ban is tractability. Regulating organophosphates is already difficult; trying to ban all pesticides would be extremely difficult politically, with profound opposition by industry and agriculture. Especially if the main benefits of such a ban were based on insect suffering.

I think there's a middle way though, following the lead of the wild animal welfare ideas I've heard. That idea would be to try to guide farmers/governments/industry to pesticides that cause less suffering as they move away from organophosphates. I would need to read and think a lot more about this to figure out my own views, but this report by the Wild Animal Initiative looks promising. Interestingly, they note a possibility that organophosphates are a (relatively) more humane class because they are faster acting. Non-pesticide based methods may be most promising.  It appears  unclear and complex though.

I'm not sure of the promise of trying to guide farmers to more humane alternatives - at face value that seems a more difficult outcome to reliably achieve than bans/de-registrations. But if work on organophosphates or pesticide-suicide prevention built networks to influence pesticide policy in LMICs, that may be a resource to layer in insect welfare efforts in the future.

Anyway, interesting comment!

Comment by Ben Stewart (BenStewart) on Cause exploration prize: organophosphate pesticides and other neurotoxicants · 2022-09-16T01:41:51.248Z · EA · GW

Hi Arno, thanks! That's a really good question. I'll say upfront that I'll be following-up with Open Phil over the coming weeks, and may have a better idea of what next steps look like then. Here's my initial stab.

I think there's three complementary approaches: 1) indirect public data/literature, 2) indirect government data, and 3) direct measurements.

  1. I had a brief look at Google Scholar and there's some interesting pieces of indirect evidence suggesting OP exposure might be a problem in Tanzania (I may be able to look into this in greater depth in the next few weeks):
    - Calista et al. (2022) appears to be an excellent and very recent source on the health effects of pesticide exposure in Tanzania, and identifies organophosphates as a concern. 
    - Boedecker et al.'s (2020) systematic review of acute unintentional pesticide poisonings suggest severely high incidence in Tanzania with 76.35% of farmers experiencing it (among the highest results in all countries considered). Thailand, which was my reference for direct measurements of OP metabolites in the urine of pregnant women, had an incidence of 36%
    - Mwevura et al. (2021) find organophosphate residues exceeding safety guidelines in soil, water, and sediment in the Rufiji River Delta in Tanzania
    - Bardner & Mcharo (2008) find that a moth species called L. meyricki, which is a pest for coffee crops, has developed general resistance to organophosphates in northern Tanzania. It appears a lot of other papers have been written about this, especially with respect to mosquito resistance as OPs have been used as vector control, e.g. mosquitoes, bedbugs. I think this is a tricky area. It might be hard to disentangle analysis of OPs used for farming vs. for vector control, and the use-practices, exposures, and cost-benefit analyses will differ. I didn't look at vector control in depth in my article, partly because the benefits seem stronger and justifying regulation seems more difficult.
    - Ngowi (2002; dissertation) finds that most pestcides used in Tanzania are organophosphates. This dissertation is long but appears thorough and very relevant, although it is quite old. 
    - Lekei et al. (2020) find that acute pesticide poisoning is common among women in Tanzania, and that organophosphates are the most important and most common type used. 
    - Lekei et al. (2014) find that organophosphates are the second most common class of pesticide used in Tanzania
    - Lekei et al. (2014b) assessed farmer's knowledge, practices, and experiences of exposure and found concerns, including widespread storage of pesticides in the home. This is also found by Manyilizu et al (2017).
    - Kapeleka (2020; thesis) found vegetables were highly contaminated with organophosphates in Tanzania.
    - Wylie et al. (2017) looked at  urinary OP metabolites in 17 pregnant women in rural Ghana, which has a lower incidence of acute unintentional pesticide poisoning according to Boedecker et al.  They found widespread exposure for  specific OP metabolites with higher levels than non-pregnant U.S. cohort - but strangely, quite low levels for non-specific metabolites (they put this down to field conditions?)
  2. FAO-STAT data is pretty coarse, but broadly suggests Tanzania is importing less hazardous pesticides over time (note however that many OP pesticides are not classified as highly hazardous). If you were collaborating with the government I would imagine that they may have much better data on import quantities and potentially usage of organophosphates specifically. 
  3. Much of the above is quite indirect. Ideally we would have good, large-sample, representative data showing levels of exposure in pregnant women in Tanzania. So far I haven't seen such a study, although I think there were some of pesticide workers in the 1990s/early 2000s. This might suggest a natural next step - fund/organise a pilot study of urinary DAPs (or other measures) in Tanzanian pregnant women from several representative communities. As well as providing direct information to guide current action, it could also serve as the beginning of a longitudinal study to  investigate the existence of a DNT effect in this context. Elikana Lekei is a Tanzanian academic whose name keeps coming up who might know a lot about this .

I hope that helps! As I said, I'll have a better idea in coming weeks of how Open Phil might want to proceed, if indeed they do. My impression of briefly looking at the literature suggests the indirect evidence for OPs being a problem in Tanzania is strong, but there's a lot more work to consider agricultural impact, regulatory environment,  pesticide alternatives, etc, as well as considering the value of getting direct evidence. I also don't know when and how approaches to government should be made, if indeed they should - but I hope others' input will clarify things in the coming months. 

Thanks for your interest and great question!

Comment by Ben Stewart (BenStewart) on Agree/disagree voting (& other new features September 2022) · 2022-09-15T00:44:01.593Z · EA · GW

Strong agree - there's plenty of posts that I think are rigorous, well-written, interesting etc., but disagree with their conclusion or general stance. It might also offer a more useful (and maybe less spicy)'sort by controversial' function, where you can see posts that are highly upvoted but torn on agreement.

Comment by Ben Stewart (BenStewart) on The (Allegedly) Best Business Books · 2022-09-12T03:36:22.519Z · EA · GW

Thanks! Charity Entrepreneurship maintain a list of books they recommend, which include a few you mention here. I'm only partway through the list and it's good and helpful, though there are some misses. 

Atomic Habits was quite practically written and concise; the Lean Start-up pretty interesting. I hated Adam Grant's 'Originals' with a passion.  Ray Dalio of Bridgewater wrote 'Principles', which has been highly recommended. The third section covering his actual management principles looks really good so far, but the second section, in which Dalio describes his philosophy of life, has some real hot garbage. 

Comment by Ben Stewart (BenStewart) on My emotional reaction to the current funding situation · 2022-09-11T22:52:20.552Z · EA · GW

This was really well-written, thanks.

Comment by Ben Stewart (BenStewart) on Cause Exploration Prizes: Announcing our prizes · 2022-09-10T07:37:22.377Z · EA · GW

Okay fair. And I just remembered something during my research that strengthens your point - there are a few papers critiquing the neurotoxic effect of organophosphates. And if memory serves all of them were authored by and/or sponsored by the pesticide industry. So there is pre-existing industry pushback!

Comment by Ben Stewart (BenStewart) on Cause Exploration Prizes: Announcing our prizes · 2022-09-09T22:06:06.787Z · EA · GW

I think you're right that industry pushback/inertia is a key consideration, and that pesticide suicides are a much more tangible harm to motivate regulation.

However, I think things are much more uncertain than you suggest. Industry have been quite on-board with the related international movement against 'highly hazardous pesticides' and the optics of 'slowly poisoning pregnant women and reducing their kids' potential' may be a compelling narrative for regulatory action.

But I really don't know - I'd be very interested to see what CPSP staff reckon.

Comment by Ben Stewart (BenStewart) on Cause Exploration Prizes: Announcing our prizes · 2022-09-09T21:57:14.319Z · EA · GW

Holy moly thank you! I'm very honoured, especially given there were so many interesting and brilliant submissions. Congratulations to all the other winners!

Comment by Ben Stewart (BenStewart) on Cause exploration prize: organophosphate pesticides and other neurotoxicants · 2022-09-09T21:49:45.839Z · EA · GW

Another worry I have is about interpersonal aggregation, which I didn't capture in the post. The 'importance' of this cause area depends on a large number of fairly mild harms (slight decrease in IQ, slight decrease in earnings, slight decrease in welfare). The large number of people affected suggests the overall magnitude is large, but I feel uncertain and queasy about suggesting it is more important than a much smaller number of serious harms (e.g. a debilitating disease or premature death).

Comment by Ben Stewart (BenStewart) on Cause exploration prize: organophosphate pesticides and other neurotoxicants · 2022-09-09T21:44:25.132Z · EA · GW

I'm very honoured that this won! Because I expect an uptick in readers for the next few days, I wanted to share an important critique that I didn't have time to incorporate properly. Tony Senanayake was overall supportive of this paper, but he was also very kind to provide these counterpoints, and to grant me permission to replicate them here:

General issue that I have: Economic empowerment is incredibly valuable to those in LMIC. This has been observed time and again by the revealed preferences of ‘beneficiaries’ in these countries. For example, we see people in India actively harming their health by moving to highly-polluted cities in India (like Delhi) in search of economic opportunities. The use of low-cost, validated, insecticides has the ability to increase yields, income and economic opportunities for farmers. These farmers often make up the plurality if not majority of the citizens in LMIC and generally are a large majority of those living in poverty in these countries. You do note that the impact on yield and livelihoods is an important component of a fuller analysis, however, I think it is something that is critical to highlight more even in this short analysis. Cause prioritization requires a fair trade-off between competing interests and if those we are looking to serve have voted by their feet that they tend to prioritize economic empowerment over health we should value that in some way. I would argue that a discounting in the BOTEC would help here. Furthermore, you may want to include in your questions the need to survey / speak with potential beneficiaries to understand their preferences as they weigh up economic empowerment and health. 

Outcome Measurement: The majority of the research you cite and the BOTEC assumptions are focused on IQ outcomes. In the development economic literature in the education space we tend to focus on student learning outcomes as measured by quality and quantity of education. The two measures that are now used as standard measures are Equivalent Years of Schooling and Learning Adjusted Years of Schooling. I would recommend adding some wording that links the IQ outcomes to EYOS or LAYS and then to measures of wellbeing (such as DALYs / QALYs / income etc). This would complete the Theory of Change that you set out. At the moment, the analysis only works through a part of the Theory of Change and leaves the reader assuming that an increase in IQ leads to welfare gains. This is a fair assumption, but not a validated one. 

Sri Lanka case study: This is a personal point for me as my family is Sri Lankan. The Sri Lankan case study is a very troublesome one. You are right that the government outlawed insecticides and pesticides and moved to ‘green agriculture’. However this is move is also seen as one of the major contributing factors to the decline in agriculture yields which precipitated the food shortages and economic crisis we are now seeing in Sri Lanka. If you are going to use Sri Lanka as an example, I think it is important to recognize these secondary consequences of policies. The unintended consequences of policies are often large and by definition uncertain and can outweigh the expected benefits.

Comment by Ben Stewart (BenStewart) on Time/Talent/Money Contributors to Existential Risk Ventures · 2022-09-06T10:28:22.003Z · EA · GW

Why are bio-related ventures called 'bio-hacking' in the posts and google forms? I understand that term to mean DIY biologists or people experimenting on their own bodies. That's pretty different to biosecurity/health security/pandemic preparedness or whatever other term for biorisk.

Comment by Ben Stewart (BenStewart) on EA & LW Forums Weekly Summary (21 Aug - 27 Aug 22’) · 2022-08-30T04:54:11.933Z · EA · GW

Thanks to you and Zoe for those Twitter lists, I've just joined and they're a great help.

Comment by Ben Stewart (BenStewart) on An Exercise in Speed-Reading: The National Security Commission on AI (NSCAI) Final Report · 2022-08-17T22:49:04.771Z · EA · GW

My impression of the literature was that reading speeds above 500-600 wpm had drastic effects on comprehension, and there being general scepticism about 'speed-reading'.

You say 'speed-reading is not simply skimming', but I think it basically is? But strategically skimming and making notes of a huge report like you did would definitely allow you to comprehend a lot (most?) of the text with massive reductions in time. It's definitely a valuable skill that should be promoted. I just think speed-reading is a bit of a false moniker.

Rapid serial visual presentation also has drawbacks because a reader can't easily pause or re-read sentences, which is often important in understanding difficult text.

Do you know of any compelling peer-reviewed evidence of speed-reading? I couldn't find any systematic reviews, but the few recent studies I found were in line with my prior view.

Comment by Ben Stewart (BenStewart) on A concern about the “evolutionary anchor” of Ajeya Cotra’s report on AI timelines. · 2022-08-16T22:43:38.745Z · EA · GW

Really interesting. I think there are connections to the extended mind thesis - where mental processes are in part constituted by the external body and world, such that cognition can't be neatly circumscribed to the brain.  This seems a deeper system than is modelled by 'computation done by neurons + training data'. 

Another complication is the relationship between ecological or environmental complexity and the evolution of intelligence. Peter Godfrey-Smith's 'Environmental Complexity and
the Evolution of Cognition' is a good read on this. Other comments on this post point to video game worlds and getting interaction by copying the evolving agent - but I think this may drastically understate the complexity of co-evolving sets of organisms in the real world.  

I think it's unlikely that developing artificial intelligence requires these wrinkles of mind extension or environmental complexity. But I interpret the evolutionary anchor argument as a generous upper bound based on what we know evolution did at least once. For that purpose, our model should probably defer to evolution's wrinkles rather than assume they're irrelevant. 


Comment by Ben Stewart (BenStewart) on Prioritisation should consider potential for ongoing evaluation alongside expected value and evidence quality · 2022-08-14T10:19:37.146Z · EA · GW

Nice. I think we could model this to see how ease/cost of evaluation interacts with other terms when assessing overall choice-worthiness. In your example the intuition sails through because A is only marginally cheaper to implement, while B is much cheaper to evaluate. I'd like to figure out  precisely when lower evaluative costs outweigh lower implementation costs, and what that depends on. 

Your post is also akin to a preference for good feedback loops when evaluating projects, which some orgs value highly. 

Comment by Ben Stewart (BenStewart) on How do independent researchers get access to resources? · 2022-08-12T08:16:00.281Z · EA · GW

And if that fails, you can usually contact the authors directly. Most academics are happy to have people interested in their work, and papers will have a corresponding author with an email address. Though obviously this method is only worth the bother if it's a really valuable paper for you.

Comment by Ben Stewart (BenStewart) on Cause area: Developmental Cognitive Neuroepidemiology · 2022-08-12T00:29:16.862Z · EA · GW

This is excellent! I wrote an entry for the competition  focused on the organophosphate pesticides you mention,  here. In that report I gesture vaguely and briefly at a wider cause area of developmental neurotoxicants. However, your proposed cause area of 'developmental cognitive neuroepidemiology' is much more systematic and ambitious. It strikes me as an excellent balance between precision of approach while remaining agnostic as to areas of focus and intervention. 

Really well done!


Comment by Ben Stewart (BenStewart) on Cause: Reducing Judicial Delay in India · 2022-08-11T04:37:38.098Z · EA · GW

Really cool, well done! I really like the explicit and quick labelling of level of evidence for each citation - I haven't seen it used outside of clinical guidelines but it seems like a nice feature given most readers are not going to look into citations.

Comment by Ben Stewart (BenStewart) on Improving incentives in research - could we topple the h-index? · 2022-08-10T09:58:54.327Z · EA · GW

I think secondary citations would be easier like you say. And you wouldn't have to stop there - once you have the citation data, you could probably do a lot of creative things analysing the resulting graphs (graphs in the mathematical sense). I expect it's where the input data is harder to reach and scrape (like whole text) that logistical worries enter.

Yeah I don't know! I'm sure there some folks who have thought about meta-science/improving science etc. that might have good ideas.

Comment by Ben Stewart (BenStewart) on New cause area: maternal morbidity · 2022-08-10T01:17:42.106Z · EA · GW

This is interesting, thanks! It's surprising that morbidity hasn't changed much despite progress on mortality, given significant overlap in their prevention/treatment. I think progress on maternal mortality could increase morbidity estimates because women are surviving with near-miss or chronic complications rather than dying. How big/real do you think this effect is?

Comment by Ben Stewart (BenStewart) on Improving incentives in research - could we topple the h-index? · 2022-08-08T10:41:29.225Z · EA · GW

Thanks, this is interesting! 2 questions and a comment:

1) Would a novelty-focused metric trade off against  replication work?

2) Would resource constraints matter for choice of metric? I'm thinking that some metrics are computationally/logistically easier to gather and maintain (e.g. pre-existing citation databases), and the cost/bother of performing textual analysis to some depth of the volumes of relevant literature might be non-negligible. 

My impression from reading some Wikipedia articles ( , , ) is that there are lots of proposals for different metrics, but a common theme of criticism is the difficulty of comparing between disciplines, where field-dependent factors are critical to a metric being meaningful/useful. If this is the case, maybe a smaller version of this project would be to pick a particularly important field to EAs, and see if targeted analysis/work can propose a more relevant metric for it. 

Comment by Ben Stewart (BenStewart) on Longtermists Should Work on AI - There is No "AI Neutral" Scenario · 2022-08-07T23:59:14.501Z · EA · GW

I don't have a deep model of AI - I mostly defer to some bodged-together aggregate of reasonable seeming approaches/people (e.g. Carlsmith/Cotra/Davidson/Karnofsky/Ord/surveys).

Comment by Ben Stewart (BenStewart) on Longtermists Should Work on AI - There is No "AI Neutral" Scenario · 2022-08-07T23:33:37.532Z · EA · GW

I'm currently involved in the UPenn tournament so can't communicate my forecasts or rationales to maintain experimental conditions, but it's at least substantially higher than 1/10,000.

And yeah, I agree complicated plans where an info-hazard makes the difference are unlikely, but info-hazards also preclude much activity and open communication about scenarios even in general. 

Comment by Ben Stewart (BenStewart) on Longtermists Should Work on AI - There is No "AI Neutral" Scenario · 2022-08-07T23:08:27.782Z · EA · GW

I'm not sure what you mean - I agree the aggregate probability of collapse is an important parameter, but I was talking about the kinds of bio-risk scenarios that simeon_c was asking for above? 
Do I understand you right that overall risk levels should be estimated/communicated even though their components might involve info-hazards? If so, I agree, and it's tricky. They'll likely be some progress on this over the next 6-12 months with Open Phil's project to quantify bio-risk, and (to some extent) the results of UPenn's hybrid forecasting/persuasion tournament on existential risks. 

Comment by Ben Stewart (BenStewart) on Longtermists Should Work on AI - There is No "AI Neutral" Scenario · 2022-08-07T22:46:40.931Z · EA · GW

Sketching specific bio-risk extinction scenarios  would likely involve substantial info-hazards.

Comment by Ben Stewart (BenStewart) on Most* small probabilities aren't pascalian · 2022-08-07T22:38:51.296Z · EA · GW

Shouldn't that last sentence be 'Pascal's mugging should not be the reason why' ?

Comment by Ben Stewart (BenStewart) on Simulated annealing, or, the importance of informal EA socialising · 2022-07-28T21:05:09.836Z · EA · GW

Really good point!

Comment by Ben Stewart (BenStewart) on Simulated annealing, or, the importance of informal EA socialising · 2022-07-28T06:32:54.170Z · EA · GW

This is also the most EA way of saying 'let's get a beer' that I could possibly come up with.

Comment by Ben Stewart (BenStewart) on Three common mistakes when naming an org or project · 2022-07-27T11:04:50.158Z · EA · GW

I've always liked the name of the Nucleic Acid Observatory - super evocative, captures the mission, memorable. It fits with point 2 here, but maybe not point 1. 

Comment by Ben Stewart (BenStewart) on Button for controversial posts · 2022-07-27T05:52:50.038Z · EA · GW

Ironically this suggestion appears to be controversial.

Comment by Ben Stewart (BenStewart) on Reducing aquatic noise as a wild animal welfare intervention · 2022-07-19T01:38:25.022Z · EA · GW

I remember a visceral introduction to this idea with this essay, describing a fall in cortisol in whale poop when shipping paused after 9/11: 

Comment by Ben Stewart (BenStewart) on Punching Utilitarians in the Face · 2022-07-14T09:44:59.574Z · EA · GW

To a certain extent, it's utilitarianism that invites these potential critiques. If a theory says that probabilities/expected value are integral to figuring out what to do, then questions looking at very large or very small probabilities/expected value is fair game.  And looking at extreme and near-extreme cases is a legitimate philosophical heuristic.

Comment by Ben Stewart (BenStewart) on Passing Up Pay · 2022-07-14T00:29:29.454Z · EA · GW

Oh true, I didn't think that bit through.  

I think there's some weaker version that still applies, for two reasons. First because the money that gets freed up may only be freed up within a specific cause area of the funder. Second because even if the freed money  goes back to the funder's general pot, it'd be distributed according to that funder's cause-area split. But if you think you've already put a lot into one cause area, maybe you'd want to push your donations more heavily in the direction of another cause area, rather than just deferring to the funder's split.

Comment by Ben Stewart (BenStewart) on Passing Up Pay · 2022-07-13T23:28:39.647Z · EA · GW

This was great and really helpful, thanks for writing it! 

This only works if there isn't some other organization you think would do more good with your money. People may overvalue their particular organization when they would be more objective in choosing among organizations at large.

Extending this thought, there's a consideration that attracts me - a kind of personal worldview diversification.  To take the example of Nucleic Acid Observatory, I might not want to go all in on pandemic prevention by both working on it and directing my counterfactual donations to it through waiving pay. I might instead want to hedge and put some of my donations into, say, animal welfare  (or neartermist global development, or what have you). 

I'm not sure if this kind of personal hedging has strong objections though. E.g. maybe it's a strategy that only makes sense if you're maximising 'the chance that I do some good', but is in tension with maximising the overall good.

Comment by Ben Stewart (BenStewart) on Vox on cash transfers vs "graduation programs" for fighting extreme poverty · 2022-07-08T00:06:24.415Z · EA · GW

Here's a link to the Banerjee paper for those without institutional access.

Comment by Ben Stewart (BenStewart) on How accurate are Open Phil's predictions? · 2022-06-17T05:34:04.340Z · EA · GW

I echo the props; what a great way to live up to your name! 

  1. How do you intend to translate  this analysis into practical change (if you think any change is warranted)?
  2. In your opinion, how do the included forecasts affect grant-making decisions? 
    My thought here is that when making a forecast about a grant, the person may not be purely playing an accuracy game - they may also be considering what the strategic/communicative impact of such forecasts are likely to be (this could be one source of the 90+% miscalibration - if I wanted to sell an idea or seem confident, I could inflate my forecast so that a particular outcome is 90+% likely).

Comment by Ben Stewart (BenStewart) on We Are Not Gods: The Geographic Critique of Impartial Progress · 2022-06-13T23:00:46.458Z · EA · GW

On context-specific prescriptions for growth, CF 'growth diagnostics'

Comment by Ben Stewart (BenStewart) on Terminate deliberation based on resilience, not certainty · 2022-06-05T23:52:02.250Z · EA · GW

This was great. This question may be too meta for its own good:

Are there plausible situations where the trend of volatility isn't stable over time? I.e. if the blue-lined envelope appears to be narrowing over deliberative effort, but then again expands wildly, or vice-versa. Call it 'chaotic volatility' for reference. 

 This might be just an extreme version of the fourth graph, but it actually seems even worse. At least in the fourth graph you might be able to recognise you're in stable or worsening volatility - in chaotic volatility you could be mistaken about what kind of epistemic situation you're in. You could think there's little further to be gained, but you're actually just before a period of narrowing volatility. Or think you're settled and confident, but with a little more deliberation a new planet swims into your ken.

One example I could think of is if someone in the general public is doing some standard career-choice agonising, doing trials, talking to people, etc. and is getting greater resilience for an option. And then on a little further reading they find 80,000 Hours and EA, and all of a sudden there's a ton more to think about and their previous resilience breaks. 

I don't know if anything action-relevant comes from considering this situation, beyond what the post already laid out. Maybe it's just trying to keep an eye out for possible  quasi-'crucial considerations' for their own choices or something. 

Comment by Ben Stewart (BenStewart) on Michael Nielsen's "Notes on effective altruism" · 2022-06-04T00:35:00.192Z · EA · GW

I loved this bit: "comfortable modernity is consistent with levels of altruistic impact and moral seriousness that we might normally associate with moral heroism"

Comment by Ben Stewart (BenStewart) on Where can I find good criticisms of EA made by non-EAs? · 2022-06-03T09:56:29.505Z · EA · GW

No worries! There's also this, very recently: 

Comment by Ben Stewart (BenStewart) on Where can I find good criticisms of EA made by non-EAs? · 2022-06-02T00:45:00.530Z · EA · GW

Although a little old (2015), this lead essay by Peter Singer in the Boston Review has a set of 11 response essays from notable academics and writers, who differ in both disciplinary background and degree of support for EA. 

I think there's probably better and more recent critiques, but it's a good spread to start with. 

Comment by Ben Stewart (BenStewart) on Stuff I buy and use: a listicle to boost your consumer surplus and productivity · 2022-06-01T23:52:03.075Z · EA · GW

Understandable, and thanks for laying your thinking. I think you're right in that we differ due to our priors. I'm probably more sceptical of  mechanistic grounding of claims in medicine, in general. This is partly due to my  experience that there's a poor correlation between how well medical interventions work in reality, and how well they should work given a seemingly strong mechanistic case. It's probably also due to a scepticism I have in general, which I haven't really justified.

Thanks for communicating so clearly - it helped me understand where my impressions were coming from, and where we disagree!

Comment by Ben Stewart (BenStewart) on Stuff I buy and use: a listicle to boost your consumer surplus and productivity · 2022-06-01T22:39:19.316Z · EA · GW

Fair, and I think the literature is really mixed, which you note elsewhere.

 For what it's worth, this recent Cochrane looked at omega-3 supplementation for depression and found a small-to-modest positive effect of Omega-3 supplementation. However, this effect was too small to be clinically significant, and the certainty was 'low to very low'. Sensitivity analysis according to whether the studies were EPA-only or predominately EPA didn't change these conclusions, except that the effect size actually got smaller compared to the effect derived from all studies.

Comment by Ben Stewart (BenStewart) on Stuff I buy and use: a listicle to boost your consumer surplus and productivity · 2022-06-01T10:19:30.291Z · EA · GW

Thanks, I like these kinds of articles. 

There’s good evidence that EPA is better than at improving mood and some evidence that DHA is better at improving cognition, and may even acutely worsen mood.

Without diving  into the literature, I don't think the evidence for the benefits of omega-3 or its subtypes is 'good' - at least not in the sources you listed or in a brief look at  Wikipedia, a skim of relevant Cochrane's, and a brief Medline search.  The evidence for benefits of supplementation in the general public is poor,  and the evidence for benefits in vegetarians/vegans is basically absent. 

It's reasonable for vegetarian/vegans to supplement it, but we should understand that doing so wouldn't be justified by  evidence - it's just appears to be very low risk and makes some sense on mechanistic grounds. 

Comment by Ben Stewart (BenStewart) on On being ambitious: failing successfully & less unnecessarily · 2022-05-26T22:30:05.770Z · EA · GW

Fantastic post Luke, thank you for writing it! 

7. We need safety nets (or we might not catch people when they fall)

One example of this kind of thing I admire is Charity Entrepreneurship's support of incubatees after the program. Even if a participant doesn't found a charity or their proposal is rejected,  CE gives them career guidance and connects them to jobs at other charities and organisations in their network, and can possibly extend the provision of a stipend. 

Comment by Ben Stewart (BenStewart) on Rational predictions often update predictably* · 2022-05-16T20:51:46.097Z · EA · GW

Bottom line up front