Does anyone know of an ethnographic study of how street gangs work? I want to understand things like how many people are involved in retail drug sales, how do the adult members interact with the juveniles, what's the level of involvement of different members, do people have specialized jobs?
I administer a rapid covid testing program at my workplace. Several dozen people, relatively close quarters, lots of exposure to the general public. Currently we have everyone doing nasal swab rapid tests every day at home before work, but there are some new preprints suggesting (with uneven quality, as befits anything based on *data* from the last couple of weeks!) that saliva samples are more sensitive, at least early on. The study design here for example https://www.medrxiv.org/content/10.1101/2021.12.22.21268246v1.full.pdf is small n and not exactly comparable to typical nasal or oral rapid test sample collection, but does anyone have an opinion on this? I'm not sure how far to update. I know there were earlier papers or at least preprints about saliva, and SalivaDirect is a thing, but I don't know to apply that in the rapid-testing-for-omicron contest.
I'm reading The Culture Code by Daniel Coyle, but I have a hard time trusting it based on the wild social science studies it cites. E.g:
# Kindergartners are better at a spaghetti tower building challenge than college students.
# "Belonging cues" (like eye contact and communication patterns) are better predictors of group performance than any other factor (including IQ of group members).
# Subjects work 50% longer on a solitary puzzle if they, early in the experiment, are given a note with not-actually-helpful information from a fictional previous subject.
# If you ask a stranger "I'm so sorry about the rain. Can I borrow your phone?", you are 422% more likely to get a yes than if you just ask "Can I borrow your phone?"
# Patients admitted after suicide attempts are 50% less likely to be readmitted if they get postcards with well-wishes from the hospital.
This is just the first 25 pages. My bullshit sensor is going crazy: all of this can't all replicate, right? Is this book worth reading?
I was reading Zvi's summary of his Omicron model, and I found that his recommendations contradict a lot of what Scott has posted on ACX:
1. He recommends that we take Vitamin D as a preventative measure against Covid. Scott of course said that he doesn't believe in this recommendation
2. He thinks taking Merck's pill is mostly bad as it will help the virus mutate more, and that it shouldn't have been approved by the CDC. Scott of course has been haranguing the CDC to approve drugs faster, etc. Of course the argument can be made that in the cost-benefit analysis, consumption of Merck's pill is the rational thing to do or something. I haven't done such a cost-benefit analysis myself. But the overall impression that I got on reading Scott's writing was that the CDC should speed up approvals.
Seeing as Scott is an influential writer who is now *clearly* influencing public policy, I wonder if he'd like to address these points. In case he's wrong about his recommendations, it might result in hundreds and thousands of preventable deaths. Of course every attempt at understanding and prediction carries with it some chance of being wrong. However, addressing views that run counter to one's own is what any good Rationalist should do anyway.
If a rapid covid antigen test shows positive results in under 1 minute is this indicative of a higher viral load than a test that takes 15 minutes to show a positive?
I just saw the Matrix Resurrections and thought it was awful. Probably the worst of the franchise.
It got me thinking: Is there any way the second and third movies (Reloaded and Revolutions) could have been fixed so that they were at least 90% as good as the first movie? If yes, then what specific things should have been different?
As a child, I was told that 15% was a standard tip, and 20% was for exceptional service. Lately, I've seen these amounts creeping upward, with 15% often being the minimum selectable option for online ordering. This would make sense if food prices are increasing more slowly than general inflation, but do we have any data suggesting this is the case? If not, what's going on?
On ranked-choice vs approval voting: I buy the argument that there are more ways to "spoil" a RCV election than an approval election, and you can be more "strategic" with RCV (lying about preferences to get your preferred candidate elected). But it doesn't change the fact that if I think the thing you're being asked to do with approval voting is nonsensical. "Mark yes for all candidates you approve of and no for the rest." Approve of compared to what? I would rather vote for any candidate on the presidential ballot last year than vote for a dog to be president, should I check all the boxes? But there's one I would like more than the others, maybe I should only check that one? It's just not clear what you're even "supposed" to do! It seems that approval voting discourages strategic voting in the technical sense (lying about your preferences) in large part by shifting your strategy to where you draw the line of vote/no vote. As far as I can see there's no honest, non-strategic way to answer that question. I personally wouldn't know what to do if presented with an approval-voting ballot.
Versus with RCV (and even FPTP) there's a pretty clear interpretation of what you're "supposed" to do: write down your candidates in the order you'd prefer seeing them elected (or mark your favorite for FPTP). You don't force people to strategically decide a cutoff, which again I think is an underspecified task. Having clear instructions for non-strategic voters is a dealbreaker for any voting system in my view.
Several years ago I was working on a fantasy world strategy/simulation that while it didn't simplify the military aspects brough the focus and attention and mechanical complexity of the other parts of the simulation up to parity. I sometimes refer to the type of game I am trying to make as a DIP game. Diplomacy|Intrigue|Politics. This is because that is what differentiates it from existing games. Whether we are talking about Paradox, Matrix, Slitherine, or purely independent games of a similar type. 4X, Grand Strategy, Political Sim(think Democracy 2-4).
As of earlier this month I am back to working on it after doing various things as well as just slacking off during the pandemic. I expect to have either release or EA on Steam/GoG/Epic/Other around early December next year.
My secondary focus is on a sort of verisimilitude where you feel much more like a politician or potentate that engages in politics rather than a god king or a hive mind. The game has both characters with more detail than CK2-3 and populations with more detail tha Vicky/Imperator. Both of these entities have an Ideology that indicates how they think the world should be organized. They also have Opinions, a numerical variable, based on various modifiers, about each other character and population they are aware of.
This, should, feel much more like a real leader trying to get things done. Opinion functions like political capital to be expended in order to do things you want to do. Start an unpopular war and get war exhaustion and ideology penalties and expend resources you could have expended for other purposes. Start a popular war and gain Opinion boosts from Ideology and if your populace hates the people you are fighting or likes that you are freeing people they like from tyranny.
Maybe you want to integrate a major minority population group from conquered lands. You can allow them to serve in government jobs/bureaucracy for a boost. But you may upset others who previously had access to that privilege. Pass various edicts/laws to give them rights, similar to Imperator but more detailed and with more options. Yes Imperator is, at least until Vicky 3 comes out, the most interesting as far as population based gameplay among Paradox games. While there isn't an overarching yes/no on immigration or w/e many policies affect parts of what we call immigration in modern societies. This game is a fantasy game so it isn't quite based on the same conflicts.
You also have to manage important/powerful characters. Flattery, bribes, titles, intimidation. I made a large effort to expand from what 4X and Grand Strategy games do with Intrigue. The Conspiracy system allows you to build up human and material capital to pursue anything from a small plot to murder another character to an "Ancient Conspiracy". A decades to millennia long secret organization to topple or usurp and empire. Essentially the Conspiracy panel/menu allows you to connect what in other games would be totally discrete actions that build up to a single greater purpose.
Many of you might be familiar with the Plot and Secrets menus from Ck2-3. But here you can do far more. You can promise land or titles or money or resources. Trade favors like military assistance, share intel from your personal "Intelligence Network", and do other stuff. The regular Diplomacy menu allows you to promise things as part of treaties but that is a little simpler and treaties are generally public. A given Conspiracy generates a Secret(you can find the 2014-2015 Axioms Of Dominion wiki which describes secrets and ideologies and all these other mechanics here: https://axioms-of-dominion.fandom.com/wiki/Axioms_Of_Dominion_Wiki) for each action of the Conspiracy and connects them in a unique way. That is the penalty for the superior organizational power of Conspiracies. Members of Conspiracies can induct others, in part or in whole, as part of their efforts to complete their parts of the overall project.
Axioms is primarily a fun video game where you can actually, compared to false claims of games in the past, feel like a character in GoT or some other media based on politics. I'm really trying to give an experience that even a modern day politician can relate to. "Why don't they just do [thing the person speaking wants]!" Because they have a complex consituency with competing and sometimes contradictory desires.
Consider the Biden BBB Bill. You have to add something to appease one Senator but it pisses off another Senator. Can you weaken the provision or add some pork to please both? Putting in a lot of long-time goals of disparate groups can gain support until you run into your very own Joe Manchin. Axioms should provide a ton of situations similar to that. Although discrete actions such as Flattering or Reasoning or even casting a literal magic mind fudgery spell will obviously not be comparable to how that works in real life. But at a higher level I think it should feel pretty close. The internal politics simulation isn't quite complex enough to have legislation and legislatures and such, at least in most cases. There's no "voting" system per se. But something like Greek democracy is pretty close to possible. And Treaties and to some degree Conspiracies function similar to how real life domestic legislation functions at a high level.
Additionally a polity can adopt a standardized "Charter" to create a governing system similar to the HRE. You can create individualized vassal contracts, I think maybe CK3 added this?, but vassals will be aware of who has a better deal, although they may also know why. Both characters and populations will have an increased opinion of "governments" that are less centralized. Triumvirates vs Dictatorships vs Councils/Parliaments, etc. There's a somewhat fixed opinion bonus, modified by ideology, as you dilute power. Of course the cost is getting all the relevant deciders on board for major group actions.
I've sketched out a way for the AI to use these mechanics, perhaps not quite to the level a player could thought process wise, such that the AI could organically form an HRE-esque superstate. A lot of this relies on the scale. I'm trying to get performance to the point that a 40000 province map can work on decently recent modern PC setups. Most people might have to make do with ~4000-8000 and perhaps a little less variety in AI political organization. I personally expect that I might need to upgrade to 32gigs of RAM to get bigger maps working.
I'm curious if a lot of ACX people play more complicated strategy games, vs Civ/Endless/GalCiv stuff.
There are *a lot* of claims in this podcast about the vaccines, the FDA, the relative harm of the variants, the comparative safety of mRNA vaccines compared to traditional vaccines, the suspected number of deaths in response to vaccination, etc.
Is there any detailed analysis and refutation or support of these claims from anyone in the rationalist community? I don't see anything (besides some allusions in 2 or 3 comments) on here, and nothing on Zvi's site.
I finally watched the Black Panther 2018 movie. (Spoilers follow.)
As a superhero action movie, I have no objection. It's fun, and it makes about as much sense as any other superhero movie, which is perfectly okay.
As a political message that doesn't even try to be subtle... when the important characters talked about difficult life of their "brothers" in USA, it was quite hypocritical from the *in-universe* perspective, in my opinion.
Hypocritical, as in: nowhere in the movie they show any concern about the well-being of their "brothers" *in Africa*; some of which are living right next to them. They probably see the poor people right out of their windows in the skyscrapers made invisible by superior technology. They keep observing them for decades or maybe centuries (not sure how long is the history of the fictional Wakanda), and they are quite okay with the quality-of-life difference.
It's only learning about mistreatment of black people *in USA* that causes the political upheaval that is the plot of the story. Which is also kinda weird, because - depending on how long Wakanda exists - why weren't they more concerned about, you know, *slavery* in USA, in the past. (Did Wakanda perhaps participate in the slave trade, and thus gained the capital to build their empire? Because having magical unobtainium is nice, but if you aren't *selling* it, it does not really explain all the wealth and modern technology.) What exactly is so special about today?
Of course, the out-universe explanation is that the movie is made today (which makes today so special), and the target audience are Americans (which makes America so special, and the rest of the world so morally irrelevant). Of course.
But it is fun to imagine a Straussian reading of the movie, as containing *two* political messages: one quite blunt, the other subtle. The Straussian reading is that no matter how much people talk about the race, wealth actually matters much more. (Even the people saying "Black Lives Matter" are ultimately talking about *American* blacks. No one needs to say it explicitly; yet everyone clearly understands that the hashtag is not about de-worming Africa.) The fictional Wakanda is a rich country; it is natural that the inhabitants only care about what happens in other rich countries. It is embarrassing for the wealthy Wakandians living incognito in USA to be treated as second-class citizens. The poverty they see out of their skyscrapers is not relevant, because they are never forced to interact with it, unless they choose to.
Therefore, I also approve of Black Panther as a subversive movie. The hidden message is that rich people need to be treated with respect, regardless of their race. Wakanda may not exist in the real world, but there are many important and wealthy people outside of USA, and we all together must strive to make their visits in USA as pleasant as possible.
Gurgaon is a city in India. It doesn't have a municipal government, but it seems to be about as good or bad as the rest of India.
There's an interesting bit of stationary bandit theory-- bribery is less expensive because there's only one agency to bribe instead of several each collecting bribes for the same thing.
I am trying to apply EA principle in other instances like solar panel installations. Does it make sense to install solar panels in your house at high north latitude or invest to install it in other places. Does anybody here have experience to model whether there is much difference in environmental impact considering location? I am assuming tropical countries may lead to more energy production per panel; how much pollution does it offset e.g. does it replace energy production from an natural oil power plant vs coal plant etc.
I am new to this community and I read a few older posts related to underweighting of sensatory input vs priors in depression and this gave me a thought. Reduction in neurotransmitters and lethargy/fatigue associated with depression might lead the brain to become more efficient in its working to conserve energy. I suspect processing sensory inputs may be more computational intensive (given how much of our brain is devoted to visual processing) This may be leading to higher weightage towards priors. My own experiences during depression episodes makes me suspect my brain just works slower and facing difficulty in learning new things. which has always made me skeptical of the depressed people are more creative theory.
Their might be a genetic component towards how much we value priors in our decision making. I noticed that there seem to be a lot of people who have suffered depression in this community than the proportion in general population. This might be people who are analytic and value prior naturally are drawn to the rationalist community.
A question for others who suffer from serious depression/panic attacks - I want to know if my below experience is a common strategy or this is totally weird? Whenever I hit my worst point and seem to collapse; I feel a rush of warm loving in my brain - like a mother loving and comforting me. I suspect this is a kind of fail safe mechanism triggering in my brain - oxytocin levels rising to reduce my stress levels.
The study authors performed 44 autopsies and found that covid RNA was present in many different tissues and in many anatomical locations, even in subjects who had mild or asymptomatic covid cases. They also found that covid was found long after infection- as far as 230 days post-infection in one case. The authors conclude that "Our data prove that SARS-CoV-2 causes systemic infection and can persist in the body for months."
It's unclear to me if this is a cause of long covid, since the authors note that they didn't see significant inflammation outside the respiratory system. I'm pretty ill-equipped to make sense of the article in general, since my degree is in a different sort of biology. I'd appreciate it if someone with a stronger understanding of immunology could take a look.
Hey Scott! How do you conduct your research for articles? Do you usually start with primary sources and scientific articles, or by getting a feel for the consensus among well-respected experts? What helps you develop new insights or come at a known problem from a new direction?
I'm hearing rumors and anecdotes that Omicron is evading rapid antigen tests due to being less prevalent in the upper respiratory tract, leading to tests that only swab the patient's nose being less effective. E.g.;
I'm cautious about this because there was also a separate trend of claims that Omicron evaded rapid tests altogether, which turned out not to be true (people tried artificial samples with various titrations of Delta and Omicron against rapid tests, and they showed roughly equal response to each.) But here there's a reasonably plausible mechanism, given the repeated studies that Omicron preferentially replicates in the bronchi.
If anyone is still looking at this slightly stale open thread, I'd appreciate people's thoughts;
- Is this for real? Was this a thing people were saying pre-Omicron? Have any formal trials been done on whether Omicron is evading nasal swabs in particular?
- Would this also effect PCR tests, or should their enhanced sensitivity be enough to catch the virus in a nasal swab regardless?
- Afaik, almost all testing done at present is via nasal swabbing alone. Would this mean that we're undercounting Omicron cases? Significantly?
A stray thought on Georgism that has suddenly appeared to me: why does not it imply that we should tax owning matter? As far as I understood, under georgism, an owner of an iron mine does not own the iron ore they have mined - the ore is owned by the whole population, and the money that the owner gets for selling the ore is not the price of the ore, but payment for the service for extracting it from the Earth, which is considerably lower. But consider the person who has bought the ore. Owning this ore is fair in the sense that they have payed for the labour of extracting it, but unfair in the sense that component of the ore which was not produced by labour - the actual existence of iron atoms - is still owned by the country population and should be taxed. And even after they process it into steel, and use it to build a skyscraper, there is still a component to those materials and things that do have not involved labour - the existence of the matter in the first place. So under those considerations, a skyscraper should actually be taxed heavier than a parking lot. And, well, owning a spoon should also be slightly taxed. Probably this could reduce consumerism or something, I don't know. (I am not actually sure this would be a good policy, just playing with the ideas and trying to look at georgism from different angles)
Hi all! I’ve been doing some research on STI prevalence / transmission rates to inform my own policy on when to ask partners to get tested. I’m considering writing up my findings for anyone else interested. I imagine this’d be about 20 pages plus lots of citations (with a shorter summary to start). I’m not a doctor, but I’m pretty thorough. Three questions:
1) Would there be interest in reading this?
2) Are there any particular questions / topics you would most want addressed in such a document?
3) Does anyone know of existing good resources that do this? I’ve seen a couple shorter attempts at discussing STI risks numerically, but wasn’t impressed by their thoroughness. (For example, they didn’t mention the fact that HIV is more likely* to transmit in the first few weeks of having it, due to high viral load at the beginning of infection.)
and especially Vermillion for a very detailed walkthrough. I only recently found out that apparently I was not getting email notifications about comment responses, and started checking my older questions for threads under them.
Even if the first AGI's values and goals are aligned with humanity's, isn't it just a matter of time before someone somewhere else builds an AGI that is misaligned? It's like nuclear power plants. As hard as you try to build them to be safe, eventually there will be a meltdown somewhere.
I didn't realize getting the flu did so much to increase stroke and heart attack risk.
The article mentions ventilation and filtration improvements to limit covid, but doesn't mention researching better treatments for people who get sick.
I have no idea what the author means by ""fairly robust death reporting in the US". In general, autopsies aren't done, and I don't know whether a full autopsy is needed to identify flu deaths.
To be fair, I don't think an autopsy could tell whether a heart attack or stroke which was an aftereffect of flu could be easily identified by any means.
The disappearance of classified threads leaves me no other place to post this, and I'm getting desperate - please delete if inappropriate.
I made a game about reasoning and fallacies, and I feel like smart, educated people interested in that or adjacent fields (logic, psychology, epistomology, statistics, ethics, plus debating and teaching) are the natural target audience, and this suggests to me that ACX is a place that might appreciate it.
Question to the AC10 hive mind: is there a way to meaningfully test the following long shot theory?
*Observation:* the vaccine denier movement seems to have brought up a wave of seriously irrational behaviour in disturbingly large numbers of people, especially in persons one would normally have considered to be more or less reasonable. In my circle of friends alone, a significant number of persons has gone off the rails in a manner that I would never have considered to be possible: with even some university-educated persons of good professional standing now believing random shit from the internet, if it just confirms their more or less bizarre pre-held convictions. Now one's own experience is of course strongly biased - but I have now heard similar stories from too many other friends that didn't got down the rabbit hole themselves to dismiss this as pure chance anymore.
*Theory:* in German, if everyone freaks out at the same time for non-obvious reasons, there is the flippant saying "es muss etwas im Trinkwasser sein" - "there has to be some contamination in the drinking water". And at first glance, that sort of thing is of course far too long a shot to be even remotely considered as a root cause of all this irrationality.
But what if there is, in fact, something in the environment right now that is causing an increased incidence of irrational behaviour in a susceptible part of the population?
What if we are seeing a version 2.0 of the low-level lead contamination that arguably caused a spike in aggressive behaviour across many nations (the "lead-crime hypothesis", https://en.wikipedia.org/wiki/Lead–crime_hypothesis).
Candidates for this would be ubiquitous chemicals that are fairly new to the environment, for instance some new plastic softening agent that has only been applied on an industrial scale in the past 15 years or so (to replace PCBs, which are deemed harmful). Not that I am actually saying that a plastic softener is to blame here - absolutely not. I'm just using those as an example of "ubiquitous to the point no one even realises that they are there anymore", and "fairly new to the environment".
*Question:* are there any statistical techniques that could be meaningfully employed to search for correlations that might uncover such a long shot chemical agent? Or is this something that would remain buried in statistical noise, even if something of the sort is playing a role?
I've read about non-specific effects of vaccines recently (https://www.bbc.com/future/article/20200915-the-mystery-of-why-some-vaccines-are-doubly-beneficial). According to that, there is benefit in training your immune system with vaccines beyond immunity against that specific illness, but also in training with illnesses themselves ("those who have been naturally infected by pathogens like measles, and lived, have better long-term survival prospects than those who were never infected"). This is something I have heard anti-vaxxers say in person, and always assumed it's either untrue or the effect is not worth the suffering of illness.
Now I'm curious whether my current strategy of avoiding any infection, not just corona, isn't doing my immune system a disservice. I'm staying away from people that have a cold even when they tested negative for Covid. My brother brought his two children to our parent's wedding this summer and I was furious when I found out that's where I got the cold from that robbed me of my only week this year. He knew they were sniffling, he brought them anyway.
His point is that if he can't bring sniffling kids, he can't bring them anywhere about half the time, because they are midly sick basically constantly. His wife, as well as a friend, both of whom are kindergardners, say that the first couple of years in the job, you are basically mildly sick all the time, and then your immune system steps up and you only rarely get the bugs the kids freely trade.
Is my policy of isolation (covid aside) a bad one? Should I be more willing to go through the headaches, the fatigue, the fever and the sore throat of common colds, for some sort of beneficial side-effect of a well-trained immune system? I've always considered coming to work or visiting friends while sick a form of physical assault (I'm in Germany, we all have insurance and sick days etc here). Opinions?
Not sure if here are mother's and father's of girls reading. If so I would like to hear your thoughts about your way of preparing, teaching, educating your soon teenage girl(s) about the potential bad sides if sexuality. I was listening to David Buss discussing parts of his new book about sexual differences in mating strategies https://www.jordanharbinger.com/david-buss-when-men-behave-badly/ (full show notes at the bottom of the long page)
Starting at roughly 21:14 minutes it's about 'womans signaling of being open for a one night stand's and resulting potential situations and confounding variables a girl or woman could get a victim of sexual violence. Sexy dressing seems to be an important factor. They make sure that everybody should be allowed to wear what he or she likes in an ideal world and don't want to conduct victim blaming. But the real world is like it is.
Do you have any advice, book, video, how to balance the right of ones daughter to dress sexy, dance sexy, have fun - and scaring her about the ultra dangerous men in the world?
Has any of you guys heard of any data about whether the immunity after COVID vaccine boosters decays any less quickly than after the primary vaccination cycle? In Turkey and in the Emirates several million people had been boosted by July, so they might probably kinda sorta be able to tell by now.
Why exactly raising hospital capacity in response to COVID waves is currently universally treated as a complete, absolute impossibility?
The calls to "flatten the curve" in March 2020 were explicitly based on the idea that while the curve is being flattened, this buys time to increase healthcare capacity. In fact, when you check out the wikipedia article "Flattening the curve", it does say in the very second paragraph: "A complementary measure is to increase health care capacity, to "raise the line".[4] As described in an article in The Nation, "preventing a health care system from being overwhelmed requires a society to do two things: 'flatten the curve'—that is, slow the rate of infection so there aren't too many cases that need hospitalization at one time—and 'raise the line'—that is, boost the hospital system's capacity to treat large numbers of patients."". And there is a nice friendly animation of the curve flattening that includes the words "increase healthcare capacity".
Nowadays though if you try to mention increasing healthcare capacity and specifically the number of ICU beds, which has always been the main issue, you always get the canned response: "this is impossible because we don't have enough personnel and it takes many years to train new ICU personnel".
But why was it widely seen as possible in March 2020? The issue wasn't really any different back then, it still was about ICU beds. Well, at first there was of course some inordinate attention to ventilators (which soon turned out not to be a big deal) but presumably you still need ICU personnel to operate these ventilators, so this does not change anything.
Here in Finland there are apparently some 250-300 ICU beds in the country of 5.5 million, and in March 2020 our health authorities claimed that this could reasonably be spun up to 1000 beds at least, and probably more. Nowadays apparently 50 ICU beds used by COVID patients is dire enough that wide-ranging restrictions for vaccinated and unvaccinated alike began again. I understand that "1000 beds" number applies when much if not most of other healthcare is winded down, just to weather the crisis. But if we're having a crisis again, why doesn't this logic apply anymore? Surely with most of the country vaccinated we still won't need anywhere close to 1000 beds, but having such a limit at 50, in a Nordic country which is widely seen as one of the most successful and functional countries in the world, sounds really rather ridiculous.
Apparently over these two years hospital capacity was in fact downsized in at least some places (at least not in Finland though), and vaccine mandates caused some nurses to quit some other places. I don't believe either of these can be a huge factor. I don't, however, understand, why no one tried, you know, PAYING COVID personnel consistently and significantly more, like 2x more than you'd normally get paid in the same position. Surely at the very least there must be a significant number of people trained as ICU nurses who are employed elsewhere, but could return to ICUs with large enough material stimulus. It's not like governments were shy about getting in debt over these two years.
But even that aside, what exactly makes it a complete, absolute impossibility to train new personnel to treat this one, specific illness under known, specific protocols? I've seem claims that it takes 5+ years to train an ICU nurse. This much time is enough to get a university degree, and getting a university degree also presupposes you need to learn to do research by yourself, write papers etc. None of that is necessary for nurses.
I find it extremely hard to believe that in 2 years it is not possible to train people who could do useful work in COVID ICUs. Admittedly with how the pandemic developed it would have been difficult for anyone in spring 2020 to predict we would actually need that; the first wave went down to near-zero in much of the world by summer and many thought (including me) that was it; then we were waiting for approval and then rollout of vaccines, and many thought (including me) that when enough people are vaccinated, this would be it. By now it seems clear that unless we're really lucky and Omicron wave is the last one (and at this point I don't really believe it anymore), this is going continue for years and we do in fact need to expand healthcare capacity. (Actually some hospitals here in Finland indeed announced that they are going to do that, if not by especially much and not especially quickly.)
But even so, why would it be impossible to train people in some months to at least help in ICUs? Sure, they wouldn't be fully capable but surely there must be some jobs to do that demand time and effort but don't require THAT much experience and knowledge. Sure, the standard of care would be reduced and more people might not survive ICU, but isn't that a crisis? If ICUs are really overflowing the standard of care would be reduced anyway.
Basically what I would really like to see is some description/timetable of which particular procedures some example COVID patients underwent while in ICU, and/or which particular procedures some example ICU nurses did over a few working days in COVID ICU. As detailed as possible, with precise notes which kind of knowledge you need to do every particular procedure. Surely even if really nothing can be done, such transparency would at least answer questions for people like me. Right now we're basically told: "these people are wizards, they are doing wizard jobs, you don't and can't understand anything about wizard jobs, no one else can ever be taught to do wizard jobs (in any meaningful timeframe at least) and everything we do is meant to make sure wizards are not overloaded in their jobs, otherwise an untold catastrophe will happen". (I also have an issue with treating overloaded ICUs as an untold catastrophe, but that's perhaps an another subject to discuss.)
I think private prisons is a bad idea. But if you have private prisons, they should be paid based on how well they rehabilitate the prisoners. So the longer a released prisoner goes without committing any crime or dying, the more money the prison gets. Maybe with a maximum of 10 years or something.
The reason the prison doesn't get money if the prisoner dies, is otherwise the prison would have an incentive to make their prisoners unhealthy, so that the prisoners would die instead of committing crimes.
I'm looking for something and maybe someone here can tell me where to find it
It's sort of a site dedicated to explaining what rent is, with little graphics of squares of land with apple trees on it. I probably found it on Reddit
Can you recommend me an article on different ethics systems?
Currently utilitarian ethics makes "most sense" to me, however it has some glaring holes in it. It would be nice to have something to lean onto in ambiguous situations, and to be self-consistent in this, hence my search for different possibilities in this area.
There's a thing that comes up here in medicine-related posts, where if you pick a random medication and look at the studies it'll tend to look really convincingly good even if we know it does nothing (like homeopathy), because of various forms of selection bias, and the conclusion there is that we should be much more skeptical than we think we should be of things that came to our attention in ways that involved selection bias.
I think this should apply to a lot of covid stuff (mainly long covid, but also things like breakthrough cases, the utility of lockdowns/NPIs in areas with available vaccines). Long covid isn't the reason covid originally came to our attention, so we should be extremely skeptical of studies claiming it exists any more than homeopathy does. There's some mildly convincing-looking studies, sure, but we'd expect that even if it's completely made up.
(I also think we should have a strong bias against NPIs in areas with available vaccines, for similar reasons)
I'm playing with the notion that being competently benevolent is a talent like being good at music, and perhaps we should view it as something we shouldn't talk as though we expect everyone to be good at it.
On the other hand, some cultures are better than others at having widely distributed ability at music, but again, it would be worth looking at how it happens.
Some companies have stopped their production of fertilizer. The reason is that petroleum gas is currently expensive, which is the basis of ammonia, which is the basis for fertilizer. So production of fertilizer is no longer profitable for some of the big players. Fertilizer is already expensive right now, and in some regions, this reduces the amount of food that we can expect to be harvested in next year's spring and summer quite drastically.
Does anybody know the scope of this? My standard newspapers are very concerned and worry about famines next year. But it is a topic where it's easy to get the order of magnitude wrong, and I don't trust them to have a good overview.
Should we expect that there are some local famines in a few places in the developing world, but no more than that, and that it can be mitigated by redistribution of food, because it is globally negligible?
Or should we expect that that there is a global food shortage? If so, how bad will it be? E.g., will it "only" affect the poorest developing countries, or also industrialized ones?
There has been recent discussion on ACX about the Food and Drug Administration (FDA) in the United States. Some believe that the FDA is overly precautious due to an asymmetry in the incentives they have or just the structure of the bureaucracy. I think these criticisms are reasonable and we should probably move toward better cost-benefit analysis. I think many people would want an FDA that approves a drug only if the amount of medical benefit after approval exceeds the possible downsides.
I am sympathetic to this view, but I think we should probably go beyond cost-benefit analysis and be inclined to approve even when it causes harms which likely exceed the benefits of approval. My reasoning is that it is much worse to prevent someone from helping themselves than to allow someone to hurt themselves. Denying access to life saving medication is worse than giving access to dangerous medication, even if both result in premature death. I will use a thought experiment from my article [1]:
"Now, imagine a nice guy named Peter who helps his elderly neighbor Betty to make sure she gets her medication when she needed it and doesn't overdose. If Peter ceased helping Betty with her medication, that would be bad. However, it would be worse if Peter were to steal his neighbors medication and withhold it form her. It would be worse even if the probability of her coming into harms way on account of either the lack of medication or inability to properly take her medication was exactly the same."
I'm working on an article right now that deals, in part, with formatted jokes, I.E. jokes as they might exist in a jokebook, where they could be memorized and retold with or without embellishments, ranging from knock-knocks to more complex dirty jokes and the like.
If you have a joke that you tell or like, it might end up being helpful to me if you told me what the joke was (a link is fine, if you don't want to type it out) and specifically *why you think the joke works*.
For full disclosure, I have pretty strong prior beliefs on this and I might end up using what you say as an example of what I disagree with. Not that I can't be convinced, but that's a strong possibility on this particular subject, so bear that in mind in your time-investing choices.
Edited to add: For the sake of Scott's sanity, probably for the best if we keep it pretty tame, especially as it relates to racial humor, which I'd just as soon take a miss on here.
I just saw a new (from earlier this month) entry on the Mistakes page:
> 41: (12/6/21) In my 2014 review of The Two Income Trap, I suggested Elizabeth Warren was smart and good. Subsequent events have conclusively revealed her to be dumb and bad. ACX regrets the error.
This prompted significant curiosity among people on my Discord server about what this means and what prompted it. Scott, would you be up for elaborating?
Does anyone know of an ethnographic study of how street gangs work? I want to understand things like how many people are involved in retail drug sales, how do the adult members interact with the juveniles, what's the level of involvement of different members, do people have specialized jobs?
Does anyone understand how continuous kilns used for brickmaking work? I just discovered them, but the webpages devoted to them are too confusing.
This is the best source I've found so far: https://civilengineering-softstudies.com/40-drying-burning-of-bricks-continuous-intermittent-bulls-trench-kiln-hoffmans-kiln.html
I administer a rapid covid testing program at my workplace. Several dozen people, relatively close quarters, lots of exposure to the general public. Currently we have everyone doing nasal swab rapid tests every day at home before work, but there are some new preprints suggesting (with uneven quality, as befits anything based on *data* from the last couple of weeks!) that saliva samples are more sensitive, at least early on. The study design here for example https://www.medrxiv.org/content/10.1101/2021.12.22.21268246v1.full.pdf is small n and not exactly comparable to typical nasal or oral rapid test sample collection, but does anyone have an opinion on this? I'm not sure how far to update. I know there were earlier papers or at least preprints about saliva, and SalivaDirect is a thing, but I don't know to apply that in the rapid-testing-for-omicron contest.
I'm reading The Culture Code by Daniel Coyle, but I have a hard time trusting it based on the wild social science studies it cites. E.g:
# Kindergartners are better at a spaghetti tower building challenge than college students.
# "Belonging cues" (like eye contact and communication patterns) are better predictors of group performance than any other factor (including IQ of group members).
# Subjects work 50% longer on a solitary puzzle if they, early in the experiment, are given a note with not-actually-helpful information from a fictional previous subject.
# If you ask a stranger "I'm so sorry about the rain. Can I borrow your phone?", you are 422% more likely to get a yes than if you just ask "Can I borrow your phone?"
# Patients admitted after suicide attempts are 50% less likely to be readmitted if they get postcards with well-wishes from the hospital.
This is just the first 25 pages. My bullshit sensor is going crazy: all of this can't all replicate, right? Is this book worth reading?
I was reading Zvi's summary of his Omicron model, and I found that his recommendations contradict a lot of what Scott has posted on ACX:
1. He recommends that we take Vitamin D as a preventative measure against Covid. Scott of course said that he doesn't believe in this recommendation
2. He thinks taking Merck's pill is mostly bad as it will help the virus mutate more, and that it shouldn't have been approved by the CDC. Scott of course has been haranguing the CDC to approve drugs faster, etc. Of course the argument can be made that in the cost-benefit analysis, consumption of Merck's pill is the rational thing to do or something. I haven't done such a cost-benefit analysis myself. But the overall impression that I got on reading Scott's writing was that the CDC should speed up approvals.
Seeing as Scott is an influential writer who is now *clearly* influencing public policy, I wonder if he'd like to address these points. In case he's wrong about his recommendations, it might result in hundreds and thousands of preventable deaths. Of course every attempt at understanding and prediction carries with it some chance of being wrong. However, addressing views that run counter to one's own is what any good Rationalist should do anyway.
https://thezvi.substack.com/p/omicron-my-current-model
If a rapid covid antigen test shows positive results in under 1 minute is this indicative of a higher viral load than a test that takes 15 minutes to show a positive?
I just saw the Matrix Resurrections and thought it was awful. Probably the worst of the franchise.
It got me thinking: Is there any way the second and third movies (Reloaded and Revolutions) could have been fixed so that they were at least 90% as good as the first movie? If yes, then what specific things should have been different?
Tipping inflation:
As a child, I was told that 15% was a standard tip, and 20% was for exceptional service. Lately, I've seen these amounts creeping upward, with 15% often being the minimum selectable option for online ordering. This would make sense if food prices are increasing more slowly than general inflation, but do we have any data suggesting this is the case? If not, what's going on?
On ranked-choice vs approval voting: I buy the argument that there are more ways to "spoil" a RCV election than an approval election, and you can be more "strategic" with RCV (lying about preferences to get your preferred candidate elected). But it doesn't change the fact that if I think the thing you're being asked to do with approval voting is nonsensical. "Mark yes for all candidates you approve of and no for the rest." Approve of compared to what? I would rather vote for any candidate on the presidential ballot last year than vote for a dog to be president, should I check all the boxes? But there's one I would like more than the others, maybe I should only check that one? It's just not clear what you're even "supposed" to do! It seems that approval voting discourages strategic voting in the technical sense (lying about your preferences) in large part by shifting your strategy to where you draw the line of vote/no vote. As far as I can see there's no honest, non-strategic way to answer that question. I personally wouldn't know what to do if presented with an approval-voting ballot.
Versus with RCV (and even FPTP) there's a pretty clear interpretation of what you're "supposed" to do: write down your candidates in the order you'd prefer seeing them elected (or mark your favorite for FPTP). You don't force people to strategically decide a cutoff, which again I think is an underspecified task. Having clear instructions for non-strategic voters is a dealbreaker for any voting system in my view.
Several years ago I was working on a fantasy world strategy/simulation that while it didn't simplify the military aspects brough the focus and attention and mechanical complexity of the other parts of the simulation up to parity. I sometimes refer to the type of game I am trying to make as a DIP game. Diplomacy|Intrigue|Politics. This is because that is what differentiates it from existing games. Whether we are talking about Paradox, Matrix, Slitherine, or purely independent games of a similar type. 4X, Grand Strategy, Political Sim(think Democracy 2-4).
As of earlier this month I am back to working on it after doing various things as well as just slacking off during the pandemic. I expect to have either release or EA on Steam/GoG/Epic/Other around early December next year.
My secondary focus is on a sort of verisimilitude where you feel much more like a politician or potentate that engages in politics rather than a god king or a hive mind. The game has both characters with more detail than CK2-3 and populations with more detail tha Vicky/Imperator. Both of these entities have an Ideology that indicates how they think the world should be organized. They also have Opinions, a numerical variable, based on various modifiers, about each other character and population they are aware of.
This, should, feel much more like a real leader trying to get things done. Opinion functions like political capital to be expended in order to do things you want to do. Start an unpopular war and get war exhaustion and ideology penalties and expend resources you could have expended for other purposes. Start a popular war and gain Opinion boosts from Ideology and if your populace hates the people you are fighting or likes that you are freeing people they like from tyranny.
Maybe you want to integrate a major minority population group from conquered lands. You can allow them to serve in government jobs/bureaucracy for a boost. But you may upset others who previously had access to that privilege. Pass various edicts/laws to give them rights, similar to Imperator but more detailed and with more options. Yes Imperator is, at least until Vicky 3 comes out, the most interesting as far as population based gameplay among Paradox games. While there isn't an overarching yes/no on immigration or w/e many policies affect parts of what we call immigration in modern societies. This game is a fantasy game so it isn't quite based on the same conflicts.
You also have to manage important/powerful characters. Flattery, bribes, titles, intimidation. I made a large effort to expand from what 4X and Grand Strategy games do with Intrigue. The Conspiracy system allows you to build up human and material capital to pursue anything from a small plot to murder another character to an "Ancient Conspiracy". A decades to millennia long secret organization to topple or usurp and empire. Essentially the Conspiracy panel/menu allows you to connect what in other games would be totally discrete actions that build up to a single greater purpose.
Many of you might be familiar with the Plot and Secrets menus from Ck2-3. But here you can do far more. You can promise land or titles or money or resources. Trade favors like military assistance, share intel from your personal "Intelligence Network", and do other stuff. The regular Diplomacy menu allows you to promise things as part of treaties but that is a little simpler and treaties are generally public. A given Conspiracy generates a Secret(you can find the 2014-2015 Axioms Of Dominion wiki which describes secrets and ideologies and all these other mechanics here: https://axioms-of-dominion.fandom.com/wiki/Axioms_Of_Dominion_Wiki) for each action of the Conspiracy and connects them in a unique way. That is the penalty for the superior organizational power of Conspiracies. Members of Conspiracies can induct others, in part or in whole, as part of their efforts to complete their parts of the overall project.
Axioms is primarily a fun video game where you can actually, compared to false claims of games in the past, feel like a character in GoT or some other media based on politics. I'm really trying to give an experience that even a modern day politician can relate to. "Why don't they just do [thing the person speaking wants]!" Because they have a complex consituency with competing and sometimes contradictory desires.
Consider the Biden BBB Bill. You have to add something to appease one Senator but it pisses off another Senator. Can you weaken the provision or add some pork to please both? Putting in a lot of long-time goals of disparate groups can gain support until you run into your very own Joe Manchin. Axioms should provide a ton of situations similar to that. Although discrete actions such as Flattering or Reasoning or even casting a literal magic mind fudgery spell will obviously not be comparable to how that works in real life. But at a higher level I think it should feel pretty close. The internal politics simulation isn't quite complex enough to have legislation and legislatures and such, at least in most cases. There's no "voting" system per se. But something like Greek democracy is pretty close to possible. And Treaties and to some degree Conspiracies function similar to how real life domestic legislation functions at a high level.
Additionally a polity can adopt a standardized "Charter" to create a governing system similar to the HRE. You can create individualized vassal contracts, I think maybe CK3 added this?, but vassals will be aware of who has a better deal, although they may also know why. Both characters and populations will have an increased opinion of "governments" that are less centralized. Triumvirates vs Dictatorships vs Councils/Parliaments, etc. There's a somewhat fixed opinion bonus, modified by ideology, as you dilute power. Of course the cost is getting all the relevant deciders on board for major group actions.
I've sketched out a way for the AI to use these mechanics, perhaps not quite to the level a player could thought process wise, such that the AI could organically form an HRE-esque superstate. A lot of this relies on the scale. I'm trying to get performance to the point that a 40000 province map can work on decently recent modern PC setups. Most people might have to make do with ~4000-8000 and perhaps a little less variety in AI political organization. I personally expect that I might need to upgrade to 32gigs of RAM to get bigger maps working.
I'm curious if a lot of ACX people play more complicated strategy games, vs Civ/Endless/GalCiv stuff.
Last night I listened to the Joe Rogan podcast with Dr. Peter McCullough.
https://open.spotify.com/episode/0aZte37vtFTkYT7b0b04Qz
There are *a lot* of claims in this podcast about the vaccines, the FDA, the relative harm of the variants, the comparative safety of mRNA vaccines compared to traditional vaccines, the suspected number of deaths in response to vaccination, etc.
Is there any detailed analysis and refutation or support of these claims from anyone in the rationalist community? I don't see anything (besides some allusions in 2 or 3 comments) on here, and nothing on Zvi's site.
I finally watched the Black Panther 2018 movie. (Spoilers follow.)
As a superhero action movie, I have no objection. It's fun, and it makes about as much sense as any other superhero movie, which is perfectly okay.
As a political message that doesn't even try to be subtle... when the important characters talked about difficult life of their "brothers" in USA, it was quite hypocritical from the *in-universe* perspective, in my opinion.
Hypocritical, as in: nowhere in the movie they show any concern about the well-being of their "brothers" *in Africa*; some of which are living right next to them. They probably see the poor people right out of their windows in the skyscrapers made invisible by superior technology. They keep observing them for decades or maybe centuries (not sure how long is the history of the fictional Wakanda), and they are quite okay with the quality-of-life difference.
It's only learning about mistreatment of black people *in USA* that causes the political upheaval that is the plot of the story. Which is also kinda weird, because - depending on how long Wakanda exists - why weren't they more concerned about, you know, *slavery* in USA, in the past. (Did Wakanda perhaps participate in the slave trade, and thus gained the capital to build their empire? Because having magical unobtainium is nice, but if you aren't *selling* it, it does not really explain all the wealth and modern technology.) What exactly is so special about today?
Of course, the out-universe explanation is that the movie is made today (which makes today so special), and the target audience are Americans (which makes America so special, and the rest of the world so morally irrelevant). Of course.
But it is fun to imagine a Straussian reading of the movie, as containing *two* political messages: one quite blunt, the other subtle. The Straussian reading is that no matter how much people talk about the race, wealth actually matters much more. (Even the people saying "Black Lives Matter" are ultimately talking about *American* blacks. No one needs to say it explicitly; yet everyone clearly understands that the hashtag is not about de-worming Africa.) The fictional Wakanda is a rich country; it is natural that the inhabitants only care about what happens in other rich countries. It is embarrassing for the wealthy Wakandians living incognito in USA to be treated as second-class citizens. The poverty they see out of their skyscrapers is not relevant, because they are never forced to interact with it, unless they choose to.
Therefore, I also approve of Black Panther as a subversive movie. The hidden message is that rich people need to be treated with respect, regardless of their race. Wakanda may not exist in the real world, but there are many important and wealthy people outside of USA, and we all together must strive to make their visits in USA as pleasant as possible.
https://www.youtube.com/watch?v=uD0bR7JXI6k&ab_channel=PolyMatterzzz
Gurgaon is a city in India. It doesn't have a municipal government, but it seems to be about as good or bad as the rest of India.
There's an interesting bit of stationary bandit theory-- bribery is less expensive because there's only one agency to bribe instead of several each collecting bribes for the same thing.
Okay, ACX people: Text-based mud for ACX and/or datasecretlox and/or ACX-adjacent people!
What would it be like?
Asking for a friend--go!!
I am trying to apply EA principle in other instances like solar panel installations. Does it make sense to install solar panels in your house at high north latitude or invest to install it in other places. Does anybody here have experience to model whether there is much difference in environmental impact considering location? I am assuming tropical countries may lead to more energy production per panel; how much pollution does it offset e.g. does it replace energy production from an natural oil power plant vs coal plant etc.
I am new to this community and I read a few older posts related to underweighting of sensatory input vs priors in depression and this gave me a thought. Reduction in neurotransmitters and lethargy/fatigue associated with depression might lead the brain to become more efficient in its working to conserve energy. I suspect processing sensory inputs may be more computational intensive (given how much of our brain is devoted to visual processing) This may be leading to higher weightage towards priors. My own experiences during depression episodes makes me suspect my brain just works slower and facing difficulty in learning new things. which has always made me skeptical of the depressed people are more creative theory.
Their might be a genetic component towards how much we value priors in our decision making. I noticed that there seem to be a lot of people who have suffered depression in this community than the proportion in general population. This might be people who are analytic and value prior naturally are drawn to the rationalist community.
A question for others who suffer from serious depression/panic attacks - I want to know if my below experience is a common strategy or this is totally weird? Whenever I hit my worst point and seem to collapse; I feel a rush of warm loving in my brain - like a mother loving and comforting me. I suspect this is a kind of fail safe mechanism triggering in my brain - oxytocin levels rising to reduce my stress levels.
A recent preprint: SARS-CoV-2 infection and persistence throughout the human body and brain (https://www.researchsquare.com/article/rs-1139035/v1)
The study authors performed 44 autopsies and found that covid RNA was present in many different tissues and in many anatomical locations, even in subjects who had mild or asymptomatic covid cases. They also found that covid was found long after infection- as far as 230 days post-infection in one case. The authors conclude that "Our data prove that SARS-CoV-2 causes systemic infection and can persist in the body for months."
It's unclear to me if this is a cause of long covid, since the authors note that they didn't see significant inflammation outside the respiratory system. I'm pretty ill-equipped to make sense of the article in general, since my degree is in a different sort of biology. I'd appreciate it if someone with a stronger understanding of immunology could take a look.
Is it possible for a state that implements Georgist land value tax policies to also implement tax farming without causing a complete fiasco?
Are we still doing the even/odd thing about whether politics is allowed?
Hey Scott! How do you conduct your research for articles? Do you usually start with primary sources and scientific articles, or by getting a feel for the consensus among well-respected experts? What helps you develop new insights or come at a known problem from a new direction?
I'm hearing rumors and anecdotes that Omicron is evading rapid antigen tests due to being less prevalent in the upper respiratory tract, leading to tests that only swab the patient's nose being less effective. E.g.;
https://twitter.com/JennyRohn/status/1475410068019585027
I'm cautious about this because there was also a separate trend of claims that Omicron evaded rapid tests altogether, which turned out not to be true (people tried artificial samples with various titrations of Delta and Omicron against rapid tests, and they showed roughly equal response to each.) But here there's a reasonably plausible mechanism, given the repeated studies that Omicron preferentially replicates in the bronchi.
If anyone is still looking at this slightly stale open thread, I'd appreciate people's thoughts;
- Is this for real? Was this a thing people were saying pre-Omicron? Have any formal trials been done on whether Omicron is evading nasal swabs in particular?
- Would this also effect PCR tests, or should their enhanced sensitivity be enough to catch the virus in a nasal swab regardless?
- Afaik, almost all testing done at present is via nasal swabbing alone. Would this mean that we're undercounting Omicron cases? Significantly?
A stray thought on Georgism that has suddenly appeared to me: why does not it imply that we should tax owning matter? As far as I understood, under georgism, an owner of an iron mine does not own the iron ore they have mined - the ore is owned by the whole population, and the money that the owner gets for selling the ore is not the price of the ore, but payment for the service for extracting it from the Earth, which is considerably lower. But consider the person who has bought the ore. Owning this ore is fair in the sense that they have payed for the labour of extracting it, but unfair in the sense that component of the ore which was not produced by labour - the actual existence of iron atoms - is still owned by the country population and should be taxed. And even after they process it into steel, and use it to build a skyscraper, there is still a component to those materials and things that do have not involved labour - the existence of the matter in the first place. So under those considerations, a skyscraper should actually be taxed heavier than a parking lot. And, well, owning a spoon should also be slightly taxed. Probably this could reduce consumerism or something, I don't know. (I am not actually sure this would be a good policy, just playing with the ideas and trying to look at georgism from different angles)
I recently wrote an article about what large language models like BERT are actually looking at when they process text. Tldr it's much stupider than you'd expect https://towardsdatascience.com/what-does-transformer-self-attention-actually-look-at-5318df114ac0
Hi all! I’ve been doing some research on STI prevalence / transmission rates to inform my own policy on when to ask partners to get tested. I’m considering writing up my findings for anyone else interested. I imagine this’d be about 20 pages plus lots of citations (with a shorter summary to start). I’m not a doctor, but I’m pretty thorough. Three questions:
1) Would there be interest in reading this?
2) Are there any particular questions / topics you would most want addressed in such a document?
3) Does anyone know of existing good resources that do this? I’ve seen a couple shorter attempts at discussing STI risks numerically, but wasn’t impressed by their thoroughness. (For example, they didn’t mention the fact that HIV is more likely* to transmit in the first few weeks of having it, due to high viral load at the beginning of infection.)
*up to 26x according to https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3130067/
I want to thank everyone who responded to this question in an earlier open thread:
https://astralcodexten.substack.com/p/open-thread-165/comment/1610069
and especially Vermillion for a very detailed walkthrough. I only recently found out that apparently I was not getting email notifications about comment responses, and started checking my older questions for threads under them.
Even if the first AGI's values and goals are aligned with humanity's, isn't it just a matter of time before someone somewhere else builds an AGI that is misaligned? It's like nuclear power plants. As hard as you try to build them to be safe, eventually there will be a meltdown somewhere.
Some of my recent writing at De Novo:
https://denovo.substack.com/p/now-hiring-for-oogenesis
The job position is still open, please apply if you think you'd be a good fit!
https://denovo.substack.com/p/hhv67-mostly-harmless
HHV6 and HHV7 are common, but mostly harmless herpesviruses. HHV6 is interesting because it can integrate into germline cells and be inherited.
https://denovo.substack.com/p/the-12-waves-of-covid
A COVID carol.
https://theair.substack.com/p/what-if-the-flu-isnt-just-the-flu
I didn't realize getting the flu did so much to increase stroke and heart attack risk.
The article mentions ventilation and filtration improvements to limit covid, but doesn't mention researching better treatments for people who get sick.
I have no idea what the author means by ""fairly robust death reporting in the US". In general, autopsies aren't done, and I don't know whether a full autopsy is needed to identify flu deaths.
To be fair, I don't think an autopsy could tell whether a heart attack or stroke which was an aftereffect of flu could be easily identified by any means.
https://emedicine.medscape.com/article/1705948-overview
Very few autopsies are done, I believe because money isn't allocated for them. We know less than we should about what kills people.
The disappearance of classified threads leaves me no other place to post this, and I'm getting desperate - please delete if inappropriate.
I made a game about reasoning and fallacies, and I feel like smart, educated people interested in that or adjacent fields (logic, psychology, epistomology, statistics, ethics, plus debating and teaching) are the natural target audience, and this suggests to me that ACX is a place that might appreciate it.
There's only three days left to fund it on kickstarter, and it's going to be close - https://www.kickstarter.com/projects/fallacy/fallacy-the-game
Question to the AC10 hive mind: is there a way to meaningfully test the following long shot theory?
*Observation:* the vaccine denier movement seems to have brought up a wave of seriously irrational behaviour in disturbingly large numbers of people, especially in persons one would normally have considered to be more or less reasonable. In my circle of friends alone, a significant number of persons has gone off the rails in a manner that I would never have considered to be possible: with even some university-educated persons of good professional standing now believing random shit from the internet, if it just confirms their more or less bizarre pre-held convictions. Now one's own experience is of course strongly biased - but I have now heard similar stories from too many other friends that didn't got down the rabbit hole themselves to dismiss this as pure chance anymore.
*Theory:* in German, if everyone freaks out at the same time for non-obvious reasons, there is the flippant saying "es muss etwas im Trinkwasser sein" - "there has to be some contamination in the drinking water". And at first glance, that sort of thing is of course far too long a shot to be even remotely considered as a root cause of all this irrationality.
But what if there is, in fact, something in the environment right now that is causing an increased incidence of irrational behaviour in a susceptible part of the population?
What if we are seeing a version 2.0 of the low-level lead contamination that arguably caused a spike in aggressive behaviour across many nations (the "lead-crime hypothesis", https://en.wikipedia.org/wiki/Lead–crime_hypothesis).
Candidates for this would be ubiquitous chemicals that are fairly new to the environment, for instance some new plastic softening agent that has only been applied on an industrial scale in the past 15 years or so (to replace PCBs, which are deemed harmful). Not that I am actually saying that a plastic softener is to blame here - absolutely not. I'm just using those as an example of "ubiquitous to the point no one even realises that they are there anymore", and "fairly new to the environment".
*Question:* are there any statistical techniques that could be meaningfully employed to search for correlations that might uncover such a long shot chemical agent? Or is this something that would remain buried in statistical noise, even if something of the sort is playing a role?
I've read about non-specific effects of vaccines recently (https://www.bbc.com/future/article/20200915-the-mystery-of-why-some-vaccines-are-doubly-beneficial). According to that, there is benefit in training your immune system with vaccines beyond immunity against that specific illness, but also in training with illnesses themselves ("those who have been naturally infected by pathogens like measles, and lived, have better long-term survival prospects than those who were never infected"). This is something I have heard anti-vaxxers say in person, and always assumed it's either untrue or the effect is not worth the suffering of illness.
Now I'm curious whether my current strategy of avoiding any infection, not just corona, isn't doing my immune system a disservice. I'm staying away from people that have a cold even when they tested negative for Covid. My brother brought his two children to our parent's wedding this summer and I was furious when I found out that's where I got the cold from that robbed me of my only week this year. He knew they were sniffling, he brought them anyway.
His point is that if he can't bring sniffling kids, he can't bring them anywhere about half the time, because they are midly sick basically constantly. His wife, as well as a friend, both of whom are kindergardners, say that the first couple of years in the job, you are basically mildly sick all the time, and then your immune system steps up and you only rarely get the bugs the kids freely trade.
Is my policy of isolation (covid aside) a bad one? Should I be more willing to go through the headaches, the fatigue, the fever and the sore throat of common colds, for some sort of beneficial side-effect of a well-trained immune system? I've always considered coming to work or visiting friends while sick a form of physical assault (I'm in Germany, we all have insurance and sick days etc here). Opinions?
Not sure if here are mother's and father's of girls reading. If so I would like to hear your thoughts about your way of preparing, teaching, educating your soon teenage girl(s) about the potential bad sides if sexuality. I was listening to David Buss discussing parts of his new book about sexual differences in mating strategies https://www.jordanharbinger.com/david-buss-when-men-behave-badly/ (full show notes at the bottom of the long page)
Starting at roughly 21:14 minutes it's about 'womans signaling of being open for a one night stand's and resulting potential situations and confounding variables a girl or woman could get a victim of sexual violence. Sexy dressing seems to be an important factor. They make sure that everybody should be allowed to wear what he or she likes in an ideal world and don't want to conduct victim blaming. But the real world is like it is.
Do you have any advice, book, video, how to balance the right of ones daughter to dress sexy, dance sexy, have fun - and scaring her about the ultra dangerous men in the world?
Speaking of Christmas and A Christmas Carol, which we weren’t - what did the Cratchit family have for dinner on Christmas Day?
Has any of you guys heard of any data about whether the immunity after COVID vaccine boosters decays any less quickly than after the primary vaccination cycle? In Turkey and in the Emirates several million people had been boosted by July, so they might probably kinda sorta be able to tell by now.
Why exactly raising hospital capacity in response to COVID waves is currently universally treated as a complete, absolute impossibility?
The calls to "flatten the curve" in March 2020 were explicitly based on the idea that while the curve is being flattened, this buys time to increase healthcare capacity. In fact, when you check out the wikipedia article "Flattening the curve", it does say in the very second paragraph: "A complementary measure is to increase health care capacity, to "raise the line".[4] As described in an article in The Nation, "preventing a health care system from being overwhelmed requires a society to do two things: 'flatten the curve'—that is, slow the rate of infection so there aren't too many cases that need hospitalization at one time—and 'raise the line'—that is, boost the hospital system's capacity to treat large numbers of patients."". And there is a nice friendly animation of the curve flattening that includes the words "increase healthcare capacity".
Nowadays though if you try to mention increasing healthcare capacity and specifically the number of ICU beds, which has always been the main issue, you always get the canned response: "this is impossible because we don't have enough personnel and it takes many years to train new ICU personnel".
But why was it widely seen as possible in March 2020? The issue wasn't really any different back then, it still was about ICU beds. Well, at first there was of course some inordinate attention to ventilators (which soon turned out not to be a big deal) but presumably you still need ICU personnel to operate these ventilators, so this does not change anything.
Here in Finland there are apparently some 250-300 ICU beds in the country of 5.5 million, and in March 2020 our health authorities claimed that this could reasonably be spun up to 1000 beds at least, and probably more. Nowadays apparently 50 ICU beds used by COVID patients is dire enough that wide-ranging restrictions for vaccinated and unvaccinated alike began again. I understand that "1000 beds" number applies when much if not most of other healthcare is winded down, just to weather the crisis. But if we're having a crisis again, why doesn't this logic apply anymore? Surely with most of the country vaccinated we still won't need anywhere close to 1000 beds, but having such a limit at 50, in a Nordic country which is widely seen as one of the most successful and functional countries in the world, sounds really rather ridiculous.
Apparently over these two years hospital capacity was in fact downsized in at least some places (at least not in Finland though), and vaccine mandates caused some nurses to quit some other places. I don't believe either of these can be a huge factor. I don't, however, understand, why no one tried, you know, PAYING COVID personnel consistently and significantly more, like 2x more than you'd normally get paid in the same position. Surely at the very least there must be a significant number of people trained as ICU nurses who are employed elsewhere, but could return to ICUs with large enough material stimulus. It's not like governments were shy about getting in debt over these two years.
But even that aside, what exactly makes it a complete, absolute impossibility to train new personnel to treat this one, specific illness under known, specific protocols? I've seem claims that it takes 5+ years to train an ICU nurse. This much time is enough to get a university degree, and getting a university degree also presupposes you need to learn to do research by yourself, write papers etc. None of that is necessary for nurses.
I find it extremely hard to believe that in 2 years it is not possible to train people who could do useful work in COVID ICUs. Admittedly with how the pandemic developed it would have been difficult for anyone in spring 2020 to predict we would actually need that; the first wave went down to near-zero in much of the world by summer and many thought (including me) that was it; then we were waiting for approval and then rollout of vaccines, and many thought (including me) that when enough people are vaccinated, this would be it. By now it seems clear that unless we're really lucky and Omicron wave is the last one (and at this point I don't really believe it anymore), this is going continue for years and we do in fact need to expand healthcare capacity. (Actually some hospitals here in Finland indeed announced that they are going to do that, if not by especially much and not especially quickly.)
But even so, why would it be impossible to train people in some months to at least help in ICUs? Sure, they wouldn't be fully capable but surely there must be some jobs to do that demand time and effort but don't require THAT much experience and knowledge. Sure, the standard of care would be reduced and more people might not survive ICU, but isn't that a crisis? If ICUs are really overflowing the standard of care would be reduced anyway.
Basically what I would really like to see is some description/timetable of which particular procedures some example COVID patients underwent while in ICU, and/or which particular procedures some example ICU nurses did over a few working days in COVID ICU. As detailed as possible, with precise notes which kind of knowledge you need to do every particular procedure. Surely even if really nothing can be done, such transparency would at least answer questions for people like me. Right now we're basically told: "these people are wizards, they are doing wizard jobs, you don't and can't understand anything about wizard jobs, no one else can ever be taught to do wizard jobs (in any meaningful timeframe at least) and everything we do is meant to make sure wizards are not overloaded in their jobs, otherwise an untold catastrophe will happen". (I also have an issue with treating overloaded ICUs as an untold catastrophe, but that's perhaps an another subject to discuss.)
I think private prisons is a bad idea. But if you have private prisons, they should be paid based on how well they rehabilitate the prisoners. So the longer a released prisoner goes without committing any crime or dying, the more money the prison gets. Maybe with a maximum of 10 years or something.
The reason the prison doesn't get money if the prisoner dies, is otherwise the prison would have an incentive to make their prisoners unhealthy, so that the prisoners would die instead of committing crimes.
I'm looking for something and maybe someone here can tell me where to find it
It's sort of a site dedicated to explaining what rent is, with little graphics of squares of land with apple trees on it. I probably found it on Reddit
Does it ring a bell with anyone?
Is there anything rational about humor?
I tend to think of it as the last exit before the highway to despair.
Can you recommend me an article on different ethics systems?
Currently utilitarian ethics makes "most sense" to me, however it has some glaring holes in it. It would be nice to have something to lean onto in ambiguous situations, and to be self-consistent in this, hence my search for different possibilities in this area.
There's a thing that comes up here in medicine-related posts, where if you pick a random medication and look at the studies it'll tend to look really convincingly good even if we know it does nothing (like homeopathy), because of various forms of selection bias, and the conclusion there is that we should be much more skeptical than we think we should be of things that came to our attention in ways that involved selection bias.
I think this should apply to a lot of covid stuff (mainly long covid, but also things like breakthrough cases, the utility of lockdowns/NPIs in areas with available vaccines). Long covid isn't the reason covid originally came to our attention, so we should be extremely skeptical of studies claiming it exists any more than homeopathy does. There's some mildly convincing-looking studies, sure, but we'd expect that even if it's completely made up.
(I also think we should have a strong bias against NPIs in areas with available vaccines, for similar reasons)
Thank you to those people who have been signing up to my substack. It is much appreciated.
However, to be absolutely clear, it is not guaranteed that I will ever publish anything. And even then, you might prefer that I hadn't...
But thank you anyway :)
I'm playing with the notion that being competently benevolent is a talent like being good at music, and perhaps we should view it as something we shouldn't talk as though we expect everyone to be good at it.
On the other hand, some cultures are better than others at having widely distributed ability at music, but again, it would be worth looking at how it happens.
Some companies have stopped their production of fertilizer. The reason is that petroleum gas is currently expensive, which is the basis of ammonia, which is the basis for fertilizer. So production of fertilizer is no longer profitable for some of the big players. Fertilizer is already expensive right now, and in some regions, this reduces the amount of food that we can expect to be harvested in next year's spring and summer quite drastically.
Does anybody know the scope of this? My standard newspapers are very concerned and worry about famines next year. But it is a topic where it's easy to get the order of magnitude wrong, and I don't trust them to have a good overview.
Should we expect that there are some local famines in a few places in the developing world, but no more than that, and that it can be mitigated by redistribution of food, because it is globally negligible?
Or should we expect that that there is a global food shortage? If so, how bad will it be? E.g., will it "only" affect the poorest developing countries, or also industrialized ones?
There has been recent discussion on ACX about the Food and Drug Administration (FDA) in the United States. Some believe that the FDA is overly precautious due to an asymmetry in the incentives they have or just the structure of the bureaucracy. I think these criticisms are reasonable and we should probably move toward better cost-benefit analysis. I think many people would want an FDA that approves a drug only if the amount of medical benefit after approval exceeds the possible downsides.
I am sympathetic to this view, but I think we should probably go beyond cost-benefit analysis and be inclined to approve even when it causes harms which likely exceed the benefits of approval. My reasoning is that it is much worse to prevent someone from helping themselves than to allow someone to hurt themselves. Denying access to life saving medication is worse than giving access to dangerous medication, even if both result in premature death. I will use a thought experiment from my article [1]:
"Now, imagine a nice guy named Peter who helps his elderly neighbor Betty to make sure she gets her medication when she needed it and doesn't overdose. If Peter ceased helping Betty with her medication, that would be bad. However, it would be worse if Peter were to steal his neighbors medication and withhold it form her. It would be worse even if the probability of her coming into harms way on account of either the lack of medication or inability to properly take her medication was exactly the same."
[1] https://parrhesia.substack.com/p/witholding
On the subject of formatted jokes:
I'm working on an article right now that deals, in part, with formatted jokes, I.E. jokes as they might exist in a jokebook, where they could be memorized and retold with or without embellishments, ranging from knock-knocks to more complex dirty jokes and the like.
If you have a joke that you tell or like, it might end up being helpful to me if you told me what the joke was (a link is fine, if you don't want to type it out) and specifically *why you think the joke works*.
For full disclosure, I have pretty strong prior beliefs on this and I might end up using what you say as an example of what I disagree with. Not that I can't be convinced, but that's a strong possibility on this particular subject, so bear that in mind in your time-investing choices.
Edited to add: For the sake of Scott's sanity, probably for the best if we keep it pretty tame, especially as it relates to racial humor, which I'd just as soon take a miss on here.
I just saw a new (from earlier this month) entry on the Mistakes page:
> 41: (12/6/21) In my 2014 review of The Two Income Trap, I suggested Elizabeth Warren was smart and good. Subsequent events have conclusively revealed her to be dumb and bad. ACX regrets the error.
This prompted significant curiosity among people on my Discord server about what this means and what prompted it. Scott, would you be up for elaborating?