All of Tsunayoshi's Comments + Replies

A case for the effectiveness of protest

Thanks for the writeup! This is surely a perspective that we are missing in EA. 

I did not have time to read all of the post, so I am not sure whether you address this: The cost-effectiveness estimates of XR are ex-post, and of just one particular organization. To me it seems obvious, that there are some movements/organizations that achieve great impact through protest, it is more difficult to determine that beforehand.  

So as far as you propose funding existing projects, do you believe that the impact and behaviour of a  movement are stable?... (read more)

1James Ozden4dThank you for the kind comments! 1. I address it in some points but yes one of my assumptions is that the cost-effectiveness of the most impactful SMOs would be within one order of magnitude of XR. 2. Regarding identifying if social movement organisations will be promising beforehand, part of the research would be to try understand what internal and external factors are crucial to social movement success. For example, I think some internal factors that might determine social movement impact could be: clear governance structure, "sticky" branding and narrative, infrastructure that will allow growth of 1000x, etc. Externally it might be things like: public salience for an issue, elite support for this issue, history of activism in that country, and so on. 3. I think this is a great question which ties into my point about re good governance structures. There is definitely a degree that the focus of social movements can shift with growth so I think it's important that the demands/campaigns are predominately set in way that allows the most impactful campaigns to be focused on, rather than what everyone finds appealing. I'm not 100% sure if this was your question so feel free to clarify if not.
Donating money, buying happiness: new meta-analyses comparing the cost-effectiveness of cash transfers and psychotherapy in terms of subjective well-being

They did not have a placebo-receiving control group. 

All the other points you mentioned seem very relevant, but I somewhat disagree with the importance of a placebo control group, when it comes to estimating counterfactual impact. If the control group is assigned to standard of care, they will know they are receiving no treatment and thus not experience any placebo effects (but unlike you write, regression-to-the-mean is still expected in that group), while the treatment group experiences placebo+"real effect from treatment". This makes it difficult t... (read more)

4peter_janicki1moThx for commenting. I have to agree with you and disagree somewhat with my earlier comment. (#placebo). Actually placebo-effects are fine and if a placebo helps people: Great! And yes, getting a specific treatment effect + the placebo-effect is better (and more like in real life), than getting no treatment at all.
An update in favor of trying to make tens of billions of dollars

 

You’d also expect that class of people to be more risk-averse, since altruistic returns to money are near-linear on relevant scales at least according to some worldviews, while selfish returns are sharply diminishing (perhaps logarithmic?).

 

It's been a while since I have delved into the topic, so take this with a grain of salt: 

Because of the heavy influence of VCs who follow a hits-based model, startup founders are often forced to aim for 1B+ companies because they lost control of the board, even if they themselves would prefer the higher ... (read more)

2Mathieu Putz2moThanks for pointing this out! Hadn't known about this, though it totally makes sense in retrospect that markets would find some way of partially cancelling that inefficiency. I've added an edit to the post.
Book Review: Open Borders

You mention "It’s probably the case that the biggest harms from immigration come from people irrationally panicking about immigration, but (surprise!) people are in fact irrational.". 

From an EU-perspective, the effect seems pretty clear: After the refugee crisis 2015-2016 there have been numerous cases of populist right-wing parties gaining support or outright winning elections after running on anti-immigration platforms: to name just a few: the Lega Nord in Italy became part of the government, the FPÖ polled at their highest in 2016,  and anti-... (read more)

We're Redwood Research, we do applied alignment research, AMA

I think it is fair to say that so far alignment research is not a standard research area in academic machine learning, unlike for example model interpretability. Do you think that would be desirable, and if so what would need to happen? 

In particular, I had this toy idea of making progress legible to academic journals:  Formulating problems and metrics that are "publishing-friendly"could,  despite the problems that optimizing for flawed metrics bring,  allow researchers at regular universities to conduct work in these areas.

4Buck2moIt seems definitely good on the margin if we had ways of harnessing academia to do useful work on alignment. Two reasons for this are that 1. perhaps non-x-risk-motivated researchers would produce valuable contributions, and 2. it would mean that x-risk-motivated researchers inside academia would be less constrained and so more able to do useful work. Three versions of this: * Somehow cause academia to intrinsically care about reducing x-risk, and also ensure that the power structures in academia have a good understanding of the problem, so that its own quality control mechanisms cause academics to do useful work. I feel pretty pessimistic about the viability of convincing large swathes of academia to care about the right thing for the right reasons. Historically, basically the only way that people have ended up thinking about alignment research in a way that I’m excited about is that they spent a really long time thinking about AI x-risk and talking about it with other interested people. And so I’m not very optimistic about the first of these. * Just get academics to do useful work on specific problems that seem relevant to x-risk. For example, I’m fairly excited about some work on interpretability and some techniques for adversarial robustness. On the other hand, my sense is that EA funders have on many occasions tried to get academics to do useful work on topics of EA interest, and have generally found it quite difficult; this makes me pessimistic about this. Perhaps an analogy here is: Suppose you’re Google, and there’s some problem you need solved, and there’s an academic field that has some relevant expertise. How hard should you try to get academics in that field excited about working on the problem? Seems plausible to me that you shouldn’t try that hard--you’d be better off trying to have a higher-touch relationship where you employ researchers or make specific grants, rather than trying to conv
Has Life Gotten Better?

Looking forward to the posts, and happy to postpone further discussion to when they are published, but  to me the question and your alluded to answer has enormous implications for our ability to raise life satisfaction levels. 

Namely: very rough estimates suggest that we are now 100x-1000x richer than in the past, and our lives are in the range [good-ok], but generally not pure bliss or anything close to it. If we extend reasonable estimations for  the effect of  material circumstances on wellbeing (i.e. doubling of wealth increases sat... (read more)

Namely: very rough estimates suggest that we are now 100x-1000x richer than in the past, and our lives are in the range [good-ok], but generally not pure bliss or anything close to it. If we extend reasonable estimations for  the effect of  material circumstances on wellbeing (i.e. doubling of wealth increases satisfaction by 1 point on a 10 point scale) , we should then expect past humans to have been miserable.

I don't think we should expect past humans to have been miserable. One of the key findings in the happiness literature is the so-called ... (read more)

Decreasing populism and improving democracy, evidence-based policy, and rationality

This is a very comprehensive report, thanks for posting.

Given that education is seen as a strong predictor of populist attitudes, it is interesting  that  many  interventions listed on the demand side  seem to target highly educated people (e.g. Our World in Data,  Factfullness,  Journalism, Fact checking in general, BPB). The Youtube channel Kurzgesagt and some things Last week tonight comes up with (e.g. the wrestler John Cena warning against conspiracy theories) seem a  bit better.  You mention research how they m... (read more)

4HaukeHillebrandt4moExcellent point. I think there's a continuum going from highly educated to those that are most at risk of populism. I haven't researched this carefully but my hunch is there are actually lots of translation of civic education memes to people who are at risk of populism (not only from experts). It seems to me that on the margin, high-quality, easily -accessible information for educated people is more neglected. related citation: IQ of the top 5% better at predicting GDP - does that suggest that increasing the epistemics of the TOP 5% is better than combating fake news? Cognitive Capitalism: The Effect of Cognitive Ability on Wealth, as Mediated Through Scientific Achievement and Economic Freedom [https://www.google.com/url?q=https://journals.sagepub.com/doi/10.1177/0956797611407207&sa=D&source=editors&ust=1627407728889000&usg=AOvVaw2CvoSVS_b0h7nt2x8A7_MI] Stefan Schubert also thinks about this this sometimes: https://stefanfschubert.com/blog/2020/10/20/fake-news-fighting-can-harm-elite-debates [https://stefanfschubert.com/blog/2020/10/20/fake-news-fighting-can-harm-elite-debates] https://stefanfschubert.com/blog/2020/12/22/legitimate-epistocracy [https://stefanfschubert.com/blog/2020/12/22/legitimate-epistocracy]
2[comment deleted]4mo
Taylor Swift's "long story short" Is Actually About Effective Altruism and Longtermism (PARODY)

Chomsky publishing his new book, The Precipice,  mere months after Long Story Short clearly indicates  that he and Taylor must be closely working together.  I look forward to the surely upcoming 80000 hours joint appearance of Taylor Swift and Noam Chomsky.  

COVID: How did we do? How can we know?

But shouldn't this update our priors towards mostly being on the happy timeline, in the West as well? Given that it took Sinovac/China one year from last March to this March to scale up, and that their vaccines are easier to manufacture than mRNA vaccines,  and if we assume high investment from the start in China (so their timeline is close to optimal), it really starts to look like we could not have done much better on manufacturing (because the West does not differ strongly in available doses compared to China)? 

I.e. we could have approved a fe... (read more)

5Ghost_of_Li_Wenliang5moNot mostly happy, I think. China apparently needed a new factory, but other places didn't [https://cepi.net/news_cepi/cepi-survey-assesses-potential-covid-19-vaccine-manufacturing-capacity/] (to the tune of 3bn wasteful doses or ~12bn real ones). Also fast approval was only one prong of the fix, along with 2) an order of magnitude more investment, 3) invested much earlier, as pre-Phase I pre-purchases, 4) HCTs, and 5) pivoting away from 80%+ waste as soon as we realise we're doing that. (HCTs are still relevant here because some of the vaccines have a shelf life < 6 months, and HCTs could thus allow May-June 2020 production to dampen the second or third waves.) Half a trillion dollars should really make some dent in the known and unknown bottlenecks. Not sure how to shrink my estimate to account for the immovable remainder.
COVID: How did we do? How can we know?

I could not agree more with your sentiment, but the "We did ok" side has a point: If there was a much better policy or intervention, why was it  done by no country, and no philanthropist?  As a country, not much was stopping you a year ago to unilaterally prepurchase tons of vaccines and start manufacturing them. Getting 20 million doses manufactured early is much easier than 2 bn, you do not need to spend time coordinating with others etc., so what happened? From memory:

China only really started to vaccinate its citizens in March (but is doing i... (read more)

6ChristianKleineidam5moBecause philantrophists like Stoecker got sued and fordidden from deploying better interventions.
2Ghost_of_Li_Wenliang5moAugust 28, 2020 [https://asia.nikkei.com/Spotlight/Coronavirus/China-s-Sinovac-starts-production-of-COVID-19-vaccine-candidate] : "Production has started at a new plant in Beijing with an annual capacity of roughly 300 million doses. Sinovac has agreed to supply 40 million doses to Bio Farma, an Indonesian state-owned company, between November and March. Sinovac started building the factory in late March and finished the project in July." So this is in fact a little piece of the happy timeline.

One last guess:

My ideology-of-all-public-officials guess is pretty weak compared to an obvious alternative: simple public-choice herding at the executive level. (200 units instead of a million.)

If governments were each minimising their own reputation loss by (correctly) predicting that they wouldn't be punished for doing what everyone was doing, this could be enough to prevent ~all innovation. As much as you want safety in numbers, you doubly don't want to be the first to risk and lose. No entrainment needed, let alone intentional coordination.

(What could ... (read more)

I am also very confused. The incentives for politicians to move as fast as they could were so vast.

Besides just vaguely accusing them of lacking courage: Another possibility is a profound entrainment of world elite opinion. One globalised and very narrow Overton window for public professionals. University is the obvious place for this to propagate, but I don't really know. What is its content? "Don't be hasty"? Could a philosophical accommodation really prevent every defection? 

(There were some - Hungary vs EU on vaccines, Israel. I actually just trie... (read more)

COVID: How did we do? How can we know?

Oh, there is not a shred of doubt that the EU delayed buying the vaccines in order to lower the price, and I agree that this was a disastrous decision that led to supply delays. This is however a separate question from approving the vaccine, which is what my objection was about.

5Ghost_of_Li_Wenliang5moAh right! Yes, I misspoke; pardon.
COVID: How did we do? How can we know?

Well whatever one may think of it, the EMA had legitimate concerns, and was not merely dragging its feet for negotiation reasons as the OP implied.

COVID: How did we do? How can we know?

"The EU's sluggishness, bullying, pandering to risk aversion, and total lack of (short-term) accountability for its own decisions. AZ approval came three months after UK approval, Pfizer three weeks after. Supposedly this was mostly taken up with haggling prices down from crazy low to crazy low."

The last sentence is uncharitable and wildly inaccurate. Do you have any evidence for this? Prior to approval, the contracts had already been finalized, and anyway, it's not the EMA (Europe's FDA equivalent) that is doing the haggling.  AZ approval was delayed... (read more)

6Ghost_of_Li_Wenliang5moHere's my source [https://www.politico.eu/article/europe-coronavirus-vaccine-struggle-pfizer-biontech-astrazeneca/] , "based on dozens of interviews with diplomats, Commission officials, pharma industry representatives and national government aides". Here's another [https://www.nytimes.com/2021/03/18/opinion/coronavirus-vaccine-europe.html], and another [https://www.eurointelligence.com/column/vaccines]. The EU's priorities are revealed in the result: 25% - 45% lower prices. They actually sort of brag about it: It seems to have been a mix of understandable coordination, show of force, price haggling, and liability haggling (which is just a kind of price haggling with extra politics). I hadn't seen those leaks, thanks. Note that they did plenty to fuel vaccine hesitancy elsewhere. No opinion about attribution to the EMA vs the Commission.
6Joe Munson5moYou mean not publishing the truth? No I think that's is bad, it's more difficult to counter vaccine skepticism when organizations are not public and transparent, at least I don't see any reason to think not.
COVID: How did we do? How can we know?

"I am extremely puzzled why China or one of the other ahem non-individualist governments didn't do these." 

Even in China, politicians and scientists fear public backlash, especially considering how to Chinese eyes the virus seems much more dangerous/lethal than to Europeans, given what happened in Wuhan:

In mainland China, scientists are intrigued by the idea, but they said it would be difficult for the public to accept.

“It is difficult to do it in China. Recently there were some articles [online about HCT] and they drew a lot of criticisms,” said Zhu

... (read more)
2Ghost_of_Li_Wenliang5moSure. Challenge trials polled well [https://pubmed.ncbi.nlm.nih.gov/33334616/] in the West, but you're right that Wuhan could have been scarifying. Test of this: how did they poll in Lombardia [https://en.wikipedia.org/wiki/COVID-19_pandemic_in_Italy]? My contention is that (in the US and UK at least) bioethicists and policymakers overestimated the controversy, possibly projecting their own misgivings.
What are the 'PlayPumps' of cause prioritisation?

I like the example of the anti-overpopulation movement of the 1960s and 70s. It involved good intentions, but its predictions and fears (e.g. widespread famines) were completely unfounded from today's perspective. It also produced some very unfortunate policies in developing countries: 

"Millions of people were sterilized, often coercively, sometimes illegally, frequently in unsafe conditions, in Mexico, Bolivia, Peru, Indonesia and Bangladesh." This article seems like a good starting point.

4EricHerboso5moThere was good reason back then to believe that overpopulation was a real problem whose time would come relatively soon. If it wasn't for technological breakthroughs [https://en.wikipedia.org/wiki/Green_Revolution] with dwarf wheat and IR8 rice variants, spearheaded by Norman Borlaug [https://en.wikipedia.org/wiki/Norman_Borlaug] and others, our population would have seriously passed our ability to grow food by this point -- the so-called Malthusian trap [https://en.wikipedia.org/wiki/Malthusianism]. Using overpopulation as an example here would be akin to using something like global climate change as an example in the present, if it turns out that a technological breakthrough in the next 5-10 years completely obviates the need for us to be careful about greenhouse gas release in the future. Because of this, I don't think overpopulation as a cause area would make for the best example that you're trying to make here.
1DavidZhang5moI guess one question I would have is whether the campaigners at the time were using good reason and evidence. It's possible that the information we have now was not available to them, and it's also possible that it's a legitimate cause area (i.e. overpopulation is a real concern) even if the means (sterilisation etc.) are clearly wrong. I'm not an expert on this at all but will read up on more recent literature on overpopulation!
2DavidZhang5moThanks so much - someone else suggested China's One Child Policy and I think this or a more general point on overpopulation might be where we end up! Really great suggestion. The sterilisation stories are harrowing and I think could really bring the point home.
What are the 'PlayPumps' of cause prioritisation?

Not sure if it fits what you are looking for, but colonialism? A somewhat underappreciated aspect of it was that the belief that spreading Western civilization would make indigenous people better off, see e.g. White man's burden. Also, the Western powers were obviously very effective at colonizing. On the other hand, it somewhat lacks the part of the "play pump" example where everybody agrees that the responsible people had good intentions. Maybe it could be adapted to tales of Christian proselytizing in Africa, which would be relevant for your audience. 

1DavidZhang5moInteresting. Without reading into it, I've always assumed that Western defences of colonialism (incl. White Man's Burden) were somewhat disingenuous, i.e. defending something they knew was wrong, or was at least controversial, and the motivations were not altruistic. The ideal case is one where people are being genuinely altruistic but completely miss the mark.
How well did EA-funded biorisk organisations do on Covid?

weeatquince's is sharing a widely held view, i.e. that eradication is superior to containment in health and economic outcomes, see e.g. this analysis. The idea itself is plausible, since a successful lockdown allows complete reopening of the internal economy afterwards.    

Sample size is however small, especially when it comes to non-island countries. I only know of two non-island countries that seriously went for eradication coupled with border closures, namely Vietnam and Israel. Israel gave up at one point when cases started to rise (which is ... (read more)

5Pablo6moThanks for the clarification. I feel that this discussion is not framed correctly. Yes, successful eradication is superior to containment in health and economic outcomes. This is a pretty weak claim that lots of people can agree with who otherwise differ considerably in their policy proposals. But the original claim was that EAs and rationalists hadn't advocated for long lockdowns and border closures, and that this was relevant for retroactively assessing their performance. The plausibility of the latter claim must be evaluated by considering all the countries that implemented long lockdowns and border closures, and not just the tiny minority that were successful in attaining (near-)eradication by adopting those measures. I took a quick look at the study you shared. Their analysis compared covid deaths, GDP growth and lockdown stringency in two groups of OECD countries during the first twelve months of the pandemic, and offered this as their original contribution to the study's main thesis that countries which favored elimination had better health and economic outcomes than countries which favored mitigation (the rest of the study is a brief and unsystematic summary of some of the relevant literature). It turns out that the group which supposedly favored elimination consists of just five nations, four of which are islands and the fifth of which (South Korea) shares borders with a single county which has been completely isolated from the rest of the world for decades. Let's pause for a moment and consider how quickly this kind of evidence would have been dismissed if it had been presented in support of a politically inconvenient conclusion. Yet here it is offered, in the world's most prestigious medical journal, to establish that "elimination, not mitigation, creates best outcomes for health, the economy, and civil liberties". For what it's worth, I personally have no strong views on how the pandemic should have been handled. (My only strong meta-view is that dec
A proposal for a small inducement prize platform

Thank you for this great post. In the past I have looked for such platforms and concepts, but was unaware of the term 'inducement prize' and did not find much. 

Two extensions to the concept you presented could make it even more interesting, especially for the EA community. Firstly, rather than just requests being supplied to such a platform, offers to conduct e.g.  research  could be posted first by  qualified researchers in order to gauge interest. Secondly, there is no reason why there couldn't be several parties/individuals who pay t... (read more)

Concerning the Recent 2019-Novel Coronavirus Outbreak

For posterity, I was wrong here because I was unaware of the dispersion parameter k that is substantially higher for SARS than for Covid-19.

Non-pharmaceutical interventions in pandemic preparedness and response

Truly excellent post! 

My intuition is that research abouts NPIs  on behavioural change might be more tractable and therefore impactful than research where the endpoint is infection.  If the endpoint is infection, any study that enrolls the general population will need to have very large sample sizes, as the examples you listed illustrate. I am sure these problems can be overcome, but I assume that one reason we have not seen more of these studies is that it is infeasible to do so without larger coordination.

 While it is unfortunate and ... (read more)

Non-pharmaceutical interventions in pandemic preparedness and response

There's an additional factor: Marketing and public persuasion. It is one thing to say: Based on a theoretical model, air filters work, and a totally different thing to say: We saw that air filters cut transmission by X% . My hope would be that the certainty and the effect estimate could serve to overcome the collective inaction we saw in the pandemic (in that many people agree that e.g. air filters would probably help, but barely nobody installed them in schools). 

1James Smith8moGood point. This is similar to what I was trying to get at when talking about lack of willingness to engage in probabilistic reasoning.
[Coronavirus] Is it a good idea to meet people indoors if everyone's rapid antigen test came back negative?

[Epistemic status: This is mostly hobbyist research that I did to evaluate which tests to buy for myself]

The numbers listed by the manufacturers are not very useful, sadly. These are generally provided without a standard protocol or independent evaluation, and can be assumed to be a best case scenario in a sample of symptomatic individuals. On the other hand, as you note, the sensitivity of antigen tests increases when infectiousness is high.

 I am absolutely out of depth trying to balance these two factors, but luckily an empirical study from the UK &... (read more)

8JP Addison8moPCR tests themselves aren't that sensitive, as mentioned by atlas. I've seen estimates (1 [https://pubmed.ncbi.nlm.nih.gov/32422057/], 2 [https://www.rcpjournals.org/content/clinmedicine/20/6/e209]) of ~20% false negative. Counterbalancing that is that I assume people who have so little virus in their nose + throat to avoid a PCR test are at lower risk of spreading. But I would (very subjectively) record my microcovids as .3 x raw numbers.
4Jonas Vollmer8moThanks, this is helpful! Feel free to PM me your payment details so I can send you the $100 reward mentioned in the post.
Dutch anti-trust regulator bans pro-animal welfare chicken cartel

No, my impression is that willingness to pay is a sufficient but not necessary condition to conclude that an industry standard benefits customers. A different sufficient condition would be an assessment of the effects of the standard by the regulators in terms of welfare. I assume that is the reason why the regulators in this case carried out an analysis of the welfare benefits, because why even do so if willingness-to-pay is the only factor? 

More speculatively, I would guess that Dutch regulators also take account welfare improvements to other  humans , and would not strike down an industry standard for safe food (if the standard actually contributed to safety). 

4Cullen_OKeefe9moI haven't read the case, but under US antitrust law this case would have the same result. The reasoning would be that individual consumers WTP for animal welfare improvements is a benefit to them, but that benefit can be realized without anticompetitive harms of raised costs and reduced variety: namely, welfare-conscious consumers can pay more for chicken raised in better conditions, and welfare-indifferent consumers still have the option to buy cheaper chickens. The discussion of "welfare" as such would therefore be a bit misleading in the US context—it's a shorthand for the maximal consumer surplus in competitive market conditions, not the type of felicific calculus EAs often do.
2Cullen_OKeefe9moI wouldn't be so sure. US antitrust authorities have repeatedly struck down pro-safety anticompetitive agreements on the justification that consumers should generally be allowed to make their own price-safety tradeoffs. See my paper that Paul linked. Of course, Dutch antitrust authorities may see it differently, but European and US antitrust analysis are usually pretty harmonized.
Google's ethics is alarming

Thank you for this post. My stance is that when engaging with hot-button topics like these, we need to pay particular attention to the truthfulness and the full picture of the topic. I am afraid that your video simplifies the reasons for the dismissal of the two researchers quite a bit to "they were fired for being critical of the AI", and would benefit from giving a fuller account. I do not want to endorse any particular side here, but it seems important to mention that 

  1. Google wanted the paper to mention that  some techniques exist to mitigate t
... (read more)
Dutch anti-trust regulator bans pro-animal welfare chicken cartel

I think you might have an incorrect impression of the ruling. The agreement was not just struck down because consumers seemed to not be willing to pay for it, but also because the ACM (on top (!) of the missing willingness to pay) decided that the agreement  did not benefit consumers by the nature of the improvements (clearly, most of the benefit goes to the chickens). 

From the link: "In order to qualify for an exemption from the prohibition on cartels under the Dutch competition regime it is necessary that the benefits passed on to the consumers exceed the harm inflicted upon them under agreements."

2Paul_Christiano9moIs your impression that if customers were willing to pay for it, then that wouldn't be sufficient cause to say that it benefited customers? (Does that mean that e.g. a standard ensuring that children's food doesn't cause discomfort also can't be protected, since it benefits customers' kids rather than customers themselves?)
vaidehi_agarwalla's Shortform

There is also a quite active EA Discord server, which serves the function of "endless group discussions" fairly well, so another Slack workspace might have negligible benefits.

EA and the Possible Decline of the US: Very Rough Thoughts

[Epistemic status: Uncertain, and also not American, so this is a 3rd party perspective]

As for the likelihood of some form of collapse, to me the current trajectory of polarization in the US seems unsustainable. Nowadays, members of both parties are split about whether they consider members of the other party "a threat to their way of life"(!)  and feelings towards the other party are rapidly declining.  

I do not think that this is just a fluke, as many political scientists argue that this is driven by an ideological sorting and a creation of a "... (read more)

3Cullen_OKeefe1yFiguring out how to move politics towards the exhausted majority seems interesting. They probably care about stability a lot more than hyper-partisans do.
80k hrs #88 - Response to criticism

Hi Mark, thanks for writing this post. I only had a cursory reading of your linked paper and the 80k episode transcript, but my impression is that Tristan's main worry (as I understand it)  and your analysis are not incompatible:  

Tristan and parts of broader society fear that through the recommendation algorithm, users discover radicalizing content. According to your paper, the algorithm does not favour and might even  actively be biased against e.g conspiracy content.

 Again, I am not terribly familiar with  the whole discussion, ... (read more)

I agree you can still criticize YouTube, even if they are recommending conspiracy content less than "view-neural". My main disagreement is with the facts - Tristan us representing YouTube is as a radicalization pipeline caused be the influence of recommendations. Let's say that YouTube is more radicalizing than a no-recommendation system all-things-considered because users were sure to click on radical content whenever it appeared. In this case you would describe radicalization as a demand from users, rather than a radialization rabbit hole caused by a man... (read more)

80k hrs #88 - Response to criticism

In general I agree, but the forum guidelines do state "Polish: We'd rather see an idea presented imperfectly than not see it at all.", and this is a post explicitly billed as "response" that were invited by Rob. So if this is all the time Mark wants to spend on it, I feel it is perfectly fine to have a post that is only for people who have listened to the podcast/are aware of the debate.

9MichaelPlant1yOh, what I said wasn't a criticism, so much as a suggestion to how more people might get up to speed on what's under debate!
Introducing Probably Good: A New Career Guidance Organization

* Good can mean quality and morality: Again, I liked that. We do mean it in both ways (the advice is both attempting to be as high quality as possibly and as high as possible in moral impact, but we are working under uncertainty in both parameters).  

For what it's worth, I liked the name specifically because to me it seemed to advertise an intention of increasing a lot of readers' impact individually by a moderate amount, unlike 80000's approach where the goal is to increase  fewer readers' impact by a large amount. 

I.e. unlike Michael I like the understatement in the name, but I agree with him that it does convey understatement. 

Avoiding Munich's Mistakes: Advice for CEA and Local Groups

Will, you are right that boycotting is not the right term for the phenomenon at hand. In addition to the reason you gave, a cancellation campaign mostly involves pressuring other organizations or people to boycott somebody. Plain old boycotting is one personal's decision  to not attend a talk, cancelling is demanding to stop the talk from even happening. 

However, I think there is some truth to the point that cancel culture is not the most productive term when used in discussions over whether it is actually a bad thing, precisely because as you sa... (read more)

2willbradshaw1yYeah, I'm sympathetic to this, and I accept the symmetry you suggest. I'm not sure to what extent it applies to this post, though.
Prioritization in Science - current view

Thanks for the write-up. Regarding the issue of loss of motivation when scientists work on research they are less intrinsically interested in: 

I know of at least one large scale historical experiment which did this. In the Soviet Union, science was reorganized to investigate areas specifically expected to increase social welfare (sadly sometimes the conclusions were predetermined by party cadres). This quote from an overview article seems relevant: 

Under the Bolshevik rule, scientists lost much of their autonomy and independence but acquired more

... (read more)
3EdoArad1yThanks! This seems very relevant. I think that my guess before glancing at this would have been that the Soviet science was affected negatively by restricting pure science, but it seems to be much more complicated than I'd naively thought. I'll definitely look more deeply into this.
N-95 For All: A Covid-19 Policy Proposal

Thanks a lot for the in depth analysis, and great analysis on the efficacy of N-95 masks. 

However, I think that because of the whole politicization of mask wearing most discussion has missed a crucial point (and I have been guilty of this as well): In situations where people are ready to wear masks (shops, public transport) infection risk is not high and surgical masks are enough. In situations where  people generally do not wear masks (bars, restaurants, private meetings at home, all day at your workplace) risk is higher but willingness to wear ... (read more)

1DannyBressler1yThis is a really good point! I think you're right that the magnitude of the benefit from the program depends heavily on how many people end up choosing to use the mask, especially in situations where they are more likely to contract the disease. Individuals will ultimately make a personal decision based on trade-offs between the probability of contracting the virus, comfort, convenience, and even fashion. I also think there is significant heterogeneity in terms of how people weigh these factors. I do think that there are a significant number of people who, net of these factors, would decide that the benefits of wearing a medical-grade respirator in situations where they are more likely to contract the virus outweighs the costs. These could be seniors, people with preexisting conditions, people who don't find the respirator uncomfortable, or people who are just risk averse. I also think that there are currently significant numbers of people who would like to wear medical-grade respirators, but who are not currently able to get them. I have friends that are teachers that are required to teach in person that want a medical-grade respirator, but are not able to get one. As I noted above, there is still a shortage of respirators even for frontline medical workers (see https://www.washingtonpost.com/business/scarcity-of-raw-material-still-squeezes-n95-mask-makers/2020/09/10/94586834-f31e-11ea-8025-5d3489768ac8_story.html). I think there are probably enough people in this category, that you could make some dent in the infection rate with this policy, though how much depends on people's behavior. Also, one of the general takeaways is that, even if the benefits end up being modest (e.g. you reduce the infection rate, but not below 1 in all areas), the relative cost is so cheap that I think it's worth it to give it a shot. A few other points: -As part of the program, it would be great to do randomized control trials with different types of respirators (e.g. different
A beginner data scientist tries her hand at biosecurity

Are you sure that this is the standard way in competitions? It is absolutely correct that before the final submission, one would find the best model by fitting it on a train set and evaluating it on the test set. However, once you found a best performing model that way, there is no reason not to train the model with the best parameters on the train+test set, and submit that one. (Submission are the predictions of the model on the validation set, not the parameters of the model).  After all, more data equals better performance.  

Avoiding Munich's Mistakes: Advice for CEA and Local Groups

Jordan Peterson is probably indeed a good example. A more objective way to describe his demeanor than shamelessness is "not giving in". One major reason why he seems to be popular is his perceived willingness to stick to controversial claims. In turn that popularity is some form of protection against attempts to get him to resign from his position at the University of Toronto.

However, I think that there are significant differences between Peterson and EA's situation, so Peterson's example is not my endorsement of a "shamelessness" strategy.

Some thoughts on the effectiveness of the Fraunhofer Society
Almost all of the contract research is done for public projects, often in joint-ventures with companies. That way, most of the funding comes from public sources.

Could you please explain that further? Looking at this document, page 13, it says that almost 50% of the proceeds from contract research is from economic sources ("Wirtschaftserträge"), and only 41 percent of the contract research money comes from public sources ("EU" and "Bund/Länder"). If my reading is correct, then it would be misleading to say that &q... (read more)

2Hans Waschke-Wischedag1yYes, about 30-35% of the total funding comes from private companies. However, these are often not direct research contracts (Company pays Fraunhofer to develop x) but publicly funded projects in which Fraunhofer works together with private companies (notably KMUinnovativ for example). But this assumption is based on my observation in different teams/institutes. I could be wrong.
EA Uni Group Forecasting Tournament!

Great idea, since you asked for suggestions, here are some:

  • I have been involved in many soccer prediction tournaments at work places or among friends. A frequent problem is that as the tournament goes on, many participants clearly have no chance of winning anymore, lose interest and drop out. Having alternative/intermediate awards (that can be purely symbolical) like "best predictions in December" or "fastest learner"(for most improved predictions) could be a fun way to mitigate that.
  • Depending on the number of people who sign up but a
... (read more)
EA considerations regarding increasing political polarization

As the post does not focus on assigning blame, it seems objectively relevant to include the alt-right. Of course definitions vary over what exactly the alt-right is but at a very basic level it is indisputable that the election of Trump has increased polarization in a way that I doubt a Mitt Romney (farther away from the far right than Trump) presidency would have.

EA considerations regarding increasing political polarization
I'd be interested to hear some reactions from EAs abroad on this article's characterization of the American "far left."

To me it seems that the focus of the post is on socio-cultural issues rather than economic. To wit, Bernie Sanders may be categorzied as part of the far left, but the focus of his platform and message were economic rather than socio-cultural. Arguably, the culture war is almost entirely focused, well on culture: i.e. the number of people being cancelled over demanding lower taxes is zero.

So from my European perspecti... (read more)

3ssalbdivad1yThat makes sense. The way the article was linked on the rising "far left" seemed to imply it was a concerning trend that young people were supportive of Democratic Socialism, despite the author never elaborating on why that would be a specific risk to EA. Perhaps it would have been clearer if the risk was broken down as a set of movements or politicians that could spawn an authoritarian government, which doesn't map very well onto a left/right spectrum.
EA considerations regarding increasing political polarization

I agree with this trend being very concerning.

One aspect that was not covered much in your post is how the EA community might be affected by increased polarization. Organizations/movements getting caught up and divided over the culture war is a frequent pattern (see e.g. the New Atheists), which would at the very least probably be very bad for trust and discourse norms in the community. Given the low percentage of some minorities in EA and EA's close relationship to academia, it is very conceivable that EA could come under more scrutiny and pressure i... (read more)

effektiv-spenden.org: 2019 - Year in Review
working hypothesis that a much broader fraction of the population might be interested in donating effectively compared to becoming full fledged and dedicated EAs

N<10, but I have had some friends approach me asking, I would like to help cause X, how can I do it most effectively. This is also one of the more frequent requests we get in our local EA group.

In general, people should have a much easier time asking the question how to e.g. support climate change most effectively, rather than whether to support climate change vs. long term risk reduction. T... (read more)

effektiv-spenden.org: 2019 - Year in Review

Sebastian, those two media reports seem rather positive and you outline your cause very well. I also realize that this is something you have less control over, but both of those clips have a clear association with EA. They both talk about William MacAskill as the founder of EA, the second one has a tag calling you "Supporter of Effective Altruism", while the first one has a separate interview with an anonymous person who is being introduced and interviewed as a member of the EA community. Thus at least those two clips *did* have a strong emphasis... (read more)

3Sebastian Schwiecker1yYeah, you are right that although we didn't emphasize it some journalists asked about EA and made it part of their coverage. I don't think that this has been negative but one challenge might be not to give the wrong impression that EA is only about donating money (we are at least aware of it).
External evaluation of GiveWell's research

Related to external evaluations: 80000hours used to have a little box at the bottom of an article, indicating a score given to it by internal and external evaluators. Does anybody know, why this is not being done anymore?

Oh man, happy to have come across this. I'm a bit surprised people remember that article. I was one of the main people that set up the system, that was a while back.

I don't know specifically why it was changed. I left 80k in 2014 or so and haven't discussed this with them since. I could imagine some reasons why they stopped it though. I recommend reaching out to them if you want a better sense.

This was done when the site was a custom Ruby/Rails setup. This functionality required a fair bit of custom coding functionality to set up. Writing quality was more

... (read more)
How to promote widespread usage of high quality, reusable masks

Point 2 is fair, but do note that I did post this under question, so some degree of uncertainty is implied.


Point 3 seems like addressing a standard that is at the very least not consistently applied. I went back through all question posts of the last 6 months; this seems to be fairly similar in form to this question and does not contain a cost-benefit analysis either. And as already mentioned, it is still part of the question category.

How to promote widespread usage of high quality, reusable masks

1) I was aware of this podcast, and actually contemplated not posting because the superficial similarity between the topics might be a problem. First of all Lewis' criticism is not just slightly different, it's all about EA's discussion of cloth masks. About medical masks, which this post is about, he specifically says "medical masks are pretty good for the general population which I’ll just about lean in favor of, although all of these things are uncertain.". Furthermore I am uncertain whether one very recent expert's opinion should promote a huge shift in EA's consensus on this topic.

3HaukeHillebrandt2yI see your point, yet still think the central argument and the other points I made still stand. I also think there are more than superficial similarities between what you suggest and what Greg criticizes, but not because both proposals are about masks, but rather because both proposals are hobbyists ideas for an intervention of dubious effectiveness.
Coronavirus Research Ideas for EAs

[Epistemic status: Uncertain]

While working on covid-19 is by no means neglected, working on unconventional solutions is. I would be surprised, if there is a serious organization outside of the EA sphere evaluating just one of these proposals:

Robin Hanson has suggested both variolation, and controlled infection of essential workers.

Distribution of medical knowledge to amateurs. As most people in the third world will not have access to medically trained professionals, they will turn to local sources of treatment. Having some people armed with a bunch of y... (read more)

[Linkpost] - Mitigation versus Supression for COVID-19

Not an expert myself, but the naive calculations that I have seen with regards to herd immunity are incorrect. The precise numbers are just to illustrate the thought process.

"We need 60-70% of people to be immune, people 65 and younger make up 65 % percent of the population, so if they catch it we have achieved herd immunity to protect the elderly".

The flaw with that reasoning is that the immune people need to be essentially randomly distributed in the population. However, the elderly make up a sub population with their own distinct networks, in... (read more)

Are there any public health funding opportunities with COVID-19 that are plausibly competitive with Givewell top charities per dollar?

Fund a study to investigate the use of DIY-masks. Promote their use to at risk population, and fund add campaigns with scientists.

Although Western governments have come out against face masks, Eastern ones (China, Japan, Korea, Taiwan) are strongly in favor of people wearing face masks to protect themselves. There is evidence, that even homemade masks are good at protecting people from respiratory droplets . These could be used if the shortage in surgical masks persist.

Since in the West nobody is promoting this evidence in favour of face masks, this is ... (read more)

3matthias_samwald2yI think so as well. I have started drafting an article about this intervention. Feel free to give feedback / share with others who might have valuable expertise: Culturally acceptable DIY respiratory protection: an urgent intervention for COVID-19 mitigation in countries outside East Asia? https://docs.google.com/document/d/11HvoN43aQrx17EyuDeEKMtR_hURzBBrYfxzseOCL5JM/edit# [https://docs.google.com/document/d/11HvoN43aQrx17EyuDeEKMtR_hURzBBrYfxzseOCL5JM/edit#]
COVID-19 brief for friends and family

There will not be a vaccine soon, but anti-viral drugs are currently in an FDA approved Phase 3 trial, and from what I have heard could be both approved and available in May.

There is evidence that higher temperatures will limit the spread: Africa has so far been mostly spared, and warm places like Singapore are doing much better than Japan or South Korea.

9steve21522yThis blog post [https://www.marklwebb.com/2020/03/covid-19-epidemiology-is-useful.html] suggests (based on Google Search Trends) that other coronavirus infections have typically gone down steadily over the course of March and April. (Presumably the data is dominated by the northern hemisphere.)

Too early to have confidence on higher temperatures limiting spread IMO (although some reason to hope, certainly); cases in japan are only <2.5x higher than singapore (234 vs 102 last I saw, and IIRC it got to japan slightly earlier); surveillance and testing in African nations unlikely to be as extensive as e.g. Japan/SK; likely less volume of travel going through african nations than some of the Asian hubs.

Any response from OpenAI (or EA in general) about the Technology Review feature on OpenAI?

It's important to react with an open mind to outside criticism of EA work, and to especially engage with the strong points. Most of the responses posted here so far (including the links to tweets of other researchers) fail to do so.

Yes, the article has a much more accusing tone than content. But, the two main criticisms are actually clear and fairly reasonable, particularly given that OpenAi (as per the article) acknowledges the importance of being respected in the greater machine learning community:

1)Whatever it is that you think about the value o... (read more)

On (2), I would note that the 'hype' criticism is one that is commonly made about the claims of both a range of individual groups in AI, and about the field as a whole. Criticisms of DeepMind's claims, and IBM's (usefulness/impact of IBM Watson in health) come immediately to mind, as well as claims by a range of groups re: deployment of self-driving cars. It's also a criticism made of the field as a whole (e.g. see various of Gary Marcus, Jack Stilgoe's comments etc). This does not necessarily mean that it's untrue of OpenAI (or that OpenAI are not one of the 'hypier'), but I think it's worth noting that this is not unique to OpenAI.

Load More