Weekly Non-Covid News #1 (10/13/22)

post by Zvi · 2022-10-13T15:40:01.523Z · LW · GW · 16 comments

Contents

  Bad News
  Good News, Everyone
  People Are Trying To Destroy the Internet
  PayPal May Not Be Your Pal, You Still Must Pay
  Fun With Bayesian Updating #1: Dustin Moskovitz AMA
  Fun With Bayesian Updating #2: No One Understands Me
  Fun With Bayesian Updating #3: And The Cheaters Gonna Cheat Cheat Cheat Cheat Cheat
  Legalize It
  People Are Increasingly Worried About Nuclear Weapons
    If there’s a fatality from an offensive nuclear detonation by 2024, will the Metaculus community prediction on the question about that be >20% for at least a day during the week before the detonation?
  Of Boys and Men
None
16 comments

This is the non-Covid part of what would previously have been the weekly Covid post. About half of the content written for the post this week is being withheld for topic-level future posts, both with longer time horizons (e.g. policy roundups don’t need to be weekly and should benefit from more integration over time) and elevating worthy sub-topics to their own posts, which this week seems likely to include the colonoscopy study.

We will retain the broad categories of Bad and Good news for various short notes.

Bad News

Sysco Teamsters are on strike in three cities, and there are a lot of people cheering the strikers on mood affectation grounds because Sysco is buying up rivals, slashing staff and service and treating everyone like garbage. Is it a monopoly worth investigating? It sure sounds like it has monopoly power and is actively seeking to assemble more of it. If you keep hiking prices (no you cannot simply say ‘inflation!’) and cutting service and everyone hates you and they live in fear of you, you might be a monopoly. If it was my job to investigate such things, I would investigate.

Signal to stop letting one send SMS messages on Android phones.

Paper (via MR) claims that there is a causal effect on call center workers where they are ‘less productive’ when they report they are happier. This effect might replicate with respect to call centers in particular, I would be shocked if it replicated to workers in general (and thus my answer to Tyler’s generalized ‘are happy workers less productive?’ is ‘no, quite the opposite, and happy to bet on this.’) If this effect is real and causal, my mechanism is that call center ‘productivity’ is about getting through calls quickly whether or not that is good for the business or customer (or victim/target), and also being willing to engage in essentially hostile interactions with the target to get the desired result. There are particular professions and times and places where one wants to Drive Angry. On the other hand I would have been more productive happier at any job I have ever had.

Kamil Galeev has an interesting thread, more explicitly than usual going beyond Russia, talking about the difference between violent and non-violent entrepreneurs, and how Musk’s instinct to ‘make a deal’ with China and Russia is exactly how the non-violent creators have gotten owned and curb-stomped for almost all of history. That in Kamil’s model of the world, where someone being allowed to create value non-violently without being wiped out is weird. His version proves too much, but the lesson he is giving here is vital, that we do not know how good we and the world have it right now in this sense. We live in a world where Musk is the rogue who doesn’t play by the rules and abuses and doesn’t respect the law. He’s the bully, not the victim. While it would be even better if he didn’t do that, we should be thrilled about this level of happy problem.

I want it to be the way where they don’t gaslight us. Alas, it is not that way.

Market pricing in three years of rough winters for Europe.

The far forward contracts might be a good buy here given the uncertainty. There are ways to have access to bountiful energy and Europe and its ‘environmentalists’ seem to be strongly against them.

The rich get higher returns than others on their real estate holdings. Presumably some of this is that rich people prefer areas that are getting richer and more expensive, whereas poorer people do not. Alternatively, rich people are buying real estate as more investment relative to consumption, and can afford to buy ‘too much house’ in order to get an area with better future prospects, and thus this effect is somewhat moderated by lower effective consumption. This still seems to suggest (very correctly, I believe) that the real estate market is not efficient and that moves in prices are predictable.

Several accounts I trust think this is best described as Cartoon Network RIP.

Autism researcher looking into theory of mind lacks good theory of Autistic mind.

I may need a podcast.

Good News, Everyone

Sarah Constantin lists some good things she wants to see more, I almost entirely agree.

Terrance Tao offers worksheet on Bayesian probability. No idea if this would be helpful to someone who didn’t already ‘get it.’

Brave browser offers feature to block cookie notifications. Tempting.

David Geffen Hall in Lincoln Center NYC has gone back to 2,200 seats as originally planned, NPR claims sound is now much improved. This is NPR’s wheelhouse, so I am inclined to believe it.

New York City finds a way to figure out that trash piles are better inside containers.

Four million dollars? If that’s what it takes? Worth every penny.

Most people who are ‘missing’ do not need or want to be found.

This is the stat that jumped out most to me. These are only the 93% where the status of not missing was verifiedso a good chunk of the remaining 7% presumably also would have fallen into this category if they had been located. Although the sheer number of ‘runaways’ was also impressive:

Note the argument from inefficiency. If a missing persons notice is not carefully geographically targeted then it seems highly unlikely to be worth sharing.

Also the periodic reminder that ‘stranger danger’ is not real, when something goes wrong it is almost always someone the victim knows.

I mostly don’t cover crypto, making an exception for this purely because it is funny.

Don’t hate the player. Hate the protocol.

Also in ‘because it’s funny,’ Patrick McKenzie has a thread that includes this line.

Never change, Patrick.

Many innovative start-up pitches start off, says Paul Graham, as ‘X for Y.’ This implies someone should program a large language model to automatically generate and evaluate an endless stream of X for Y ideas and find the good ones. As a startup?

People Are Trying To Destroy the Internet

Yout legal ruling seems to be latest attempt to destroy the internet if you look at the details, featuring such winners as ‘modifying the URL to access content is illegal under 1201’ and ‘it counts as an effective protection measure if you don’t actively have a feature allowing it’ so anything that adds a feature to a website would be illegal.

The sheer rate of such rulings and warnings I now see, in large part due to following Mike Masnick who both is on top of these things and likes to point out how rulings might destroy things, implies that they are mostly overblown and/or that someone out there is dealing with them. I hope. There really are a lot of them, and he’s also warning us that another ruling could potentially destroy art by decimating fair use.

PayPal May Not Be Your Pal, You Still Must Pay

It is important to note when things still have the ability to surprise you.

Whoops. Our bad.

Musk is responding to Marcus here.

The announcement of reversal is always interesting (Reuters version):

I wonder what happens on November 4th?

I wonder how something like this gets into your policy ‘in error’?

It also is only a partial reversal, as the Volokh Conspiracy notes that there are still $2500 fines for ‘intolerance’ or ‘discriminatory’ content. One I like is ‘involve the sales of products or services identified by government agencies to have a high likelihood of being fraudulent’ even if they’re not fraudulent. The whole current policy is essentially carte blanche if someone wants to get creative.

To be fair, this isn’t entirely a new tactic, PayPal is a fan of ‘take the money and laugh.’

It is never clear how many people respond to such things.

Does anyone know an instance of PayPal taking money out of your bank account in this way, unauthorized? Would there be legal remedy if they did? As in, if they felt like it, could they literally change their terms of service again, then decide to drain the entire linked bank account, and then that’s that?

What is clear, whether or not I close the account, is that I have no intention of ever keeping a substantial PayPal account balance ever again.

The market disliked what it saw, stock down 9% despite partial reversal.

Fun With Bayesian Updating #1: Dustin Moskovitz AMA

Effective Altruism Twitter AMA by Dustin Moskovitz.

Mostly basic questions, mostly useful for insight into how Dustin personally thinks.

In terms of our disagreements, this confirmed how deeply he is committed to his effectiveness calculations, believes in institutions (e.g. he thinks world government would be an importantly good thing), doesn’t model first world or American conditions as having strong spillover effects (although he donates to the Democrats anyway, and is properly worried about 2024), and defers to certain individuals.

He mentioned he’s soured on prizes, I think unfairly given how much has been spent on them (aka not much).

This was the first one I found surprising:

Then there’s this:

I wouldn’t have him buy a giant yacht, but I say Dustin Moskovitz should go to space. It is not that expensive, and doing so would be highly educational on many levels. Worth it on that alone, also come on man, you’ve earned it.

An underappreciated point by many, you love to see it:

Love me some Zohar Atkins, insightful including into how Dustin thinks about ‘interventions’:

Fun With Bayesian Updating #2: No One Understands Me

Derek Thompson at The Atlantic asks people, what no one understands about their job. It’s a fun one. It’s both a chance both to learn rarely known things about people’s jobs, and also to see how well-calibrated people are about what other people know or expect about their jobs.

The biggest pattern is pure Robin Hanson. You think my job doing X is mostly about X and I’m free to X all day. Well, you’re wrong. It’s actually about this other less exciting thing Y, usually a mix of paperwork and meetings and other things no one likes, which is what allows me to do what little X the job as a Doer of X affords me. This came up [12] times (Aid Worker, Book Editor, Chef, Corporate Communications Executive, Data Scientist 2, Opera Singer, Pastor, Pharmacist, Postal Worker, Sailor, Screenwriter, Camp Director), of which [3] was surprising to me.

I do think a lot of people lack the knowledge being given here. Even for me, it was good to see it spelled out. I would say to these folks, some of us do get it. In some cases, I’m guessing most of us.

Then there is the flip side, where you think we all do X all day and can’t or never do Y, but actually we also do Y and it’s central so stop stereotyping and selling us short (Data Scientist 1, Debate Coach, Nurse, Real Estate Agent, Timpanist). This came up [5] times and was surprising [2] times.

There were [9] others (ER doctor, Financial Analyst, Humanitarian, Government Consultant, Grape Grower, Lobbyist 1, Lobbyist 2, Neuroscientist, Managing Director, Software Engineer, Stenographer), which were basically ‘yes this job is exactly what you think it is’ and were thus surprising to me zero times. To them, I say that actually, Everybody Knows. People are not as clueless as you think.

Fun With Bayesian Updating #3: And The Cheaters Gonna Cheat Cheat Cheat Cheat Cheat

The investigation into the poker cheating incident at Hustler Casino continues.

It does not get less weird. It does get clearer.

Even if some things are a coincidence, this does not seem like a coincidence.

So let me get this straight. After the incident, on camera, someone who is involved in the production took $15,000 chips out of Robbi’s stack. The same stack which is counted continuously on camera, so she has to know she was shorted, from the same recording that contained an incident sure to be investigated in some fashion. Then she declined to prosecute.

As someone else put it, this has very large ‘questions already answered by the shirt’ vibes to it. I can think of one very obvious explanation for these facts. I cannot think of an alternative explanation for these facts.

After I wrote that, Doug Polk weighed in, including with some new information. All of it points towards this being cheating, and offers plausible mechanisms for how it happened. He can’t come up with a plausible alternative explanation either.

Here is Garrett Adelstein (the other one in the original hand) on all this, he unsurprisingly has high confidence of cheating and details.

Why am I paying this much attention to this? I find it fascinating, it is Relevant To My Interests, I find it a great place to practice thinking about what is going on in a situation, and most of all we see a lot of great examples of people doing their best to think well and do so out loud in public, and what different people focus on and think is important.

Seems only fair, then, to give my percentage as well. I’m at something like 90% cheat at this point, due to not being an expert and thus having a bunch of model uncertainty.

Also only fair, the fishing cheater has been indicted. Excellent.

Legalize It

Biden issues full pardons for all federal cases of simple marijuana possession, moves to reclassify marijuana as a drug that is not, as it turns out, on the same level as heroin and more dangerous than fentanyl.

There are about 6,500 such cases, because most drug convictions are at the state level, so this is starting small. If governors honor Biden’s call to do this at the state level as well, then the direct effect will be much larger. If the reclassification goes through, that too would be a huge deal.

Even without those steps, this seems like a big additional step towards effective legalization of marijuana. That is not an unmitigated good. If I had to choose whether people should generally smoke more marijuana or less marijuana, I would choose less. There are times and places where it makes life better, yet in general marginal consumption is something I would expect to make life worse. Drugs that make one stupid, that are anti-motivational, that cause accidents and are often used irresponsibly, are drugs it is wise to mostly avoid, and for society to discourage.

I would say all the same things for alcohol, and also think prohibition of either of them makes a similar amount of sense. The popularity of those two proposals is also converging, straight towards being very, very unpopular.

Thus, when I announced Balsa, I put full marijuana legalization on the list of obvious wins. It was the thing on the list I was most confident would likely happen on its own. Which seems to be exactly what is happening here. Good job all around.

Now we need to study marijuana to know how to reschedule it, except of course that it being scheduled makes it much harder to do the study.

“It’s something that we constantly communicate: We really need to figure out a way of doing research with these substances,” Nora Volkow, thedirector of the National Institute on Drug Abuse, told The Health 202. Her agency is a major funder of studies on drug use and addiction, and Volkow said she’s working to find pathways to ease research into Schedule I drugs.

Either way, a final determination over how to classify marijuana could take years. The prospect is sure to ignite a flurry of lobbying and a renewed push in Congress to decriminalize the drug at the federal level.

People Are Increasingly Worried About Nuclear Weapons

Up front overall view here, so it isn’t missed: While nuclear war risk is higher than normal and highly unacceptable, in my view it remains low, and the Metaculus and prediction market estimates seem substantially too high. Please do not obsess or lose sleep over this unless you are trying to change the outcome. I do not believe any costly actions (like relocating) are justified at this time in terms of personal safety, although purchasing basic emergency supplies is a good idea even under baseline conditions so nothing wrong with doing that. If you are going to have a trigger for leaving, I believe the most aggressive reasonable trigger is a Russian nuclear detonation, and it would also be reasonable to continue to wait until there was an offensive detonation.

OK, then. Onward.

I give this week’s Good Use of Prediction Markets award to this question:

Another question in this tournament asks Will an offensive nuclear detonation cause at least 1 fatality before 2024?. If that does happen, how predictable will it be in the days beforehand? To what extent will it be a bolt from the blue?

If there’s a fatality from an offensive nuclear detonation by 2024, will the Metaculus community prediction on the question about that be >20% for at least a day during the week before the detonation?

This question conditions on Will an offensive nuclear detonation cause at least 1 fatality before 2024? resolving positively; that is, this question resolves ambiguously unless that other one resolves positively.

This question resolves positively if:

1. That other question resolves positively, and
2. For a full 24 hour period sometime in the 168 hours (i.e., 7 times 24) before the first detonation that resolves that other question positively, the Community Prediction on that other question is greater than 20%.

Then there’s the question of how high risk is now. Everyone agrees it is too damn high, because it is always too damn high and also it is clearly elevated. How far?

Here’s the current prediction of this risk, which was 7% when I grabbed it (it has since gone down, then back up to 9%). The distribution is that most people (including me) think that is high, while some think the risk is very high.

Here is Polymarket with the question that ends at the end of 2022 rather than 2023, and restricted purely to Russia and ‘in an offensive capacity’:

The new meta-question is a great question to be asking. If we are worried about whether to bug out or otherwise do something about to protect ourselves, we need to know how much warning we will have that things are likely to go south. If we used a 20% threshold on this market as our trigger, how often would it be in time?

Responses are all over the map from 2% to 90%, with the community prediction at 50% chance that Metaculus will flag the risk at 20%+ for at least 24 hours before the first detonation, which could plausibly be a point of no return where a lot of actions get impossible or far more expensive.

My evaluation is that the 50% here is low, given a good quickly-updating-on-new-predictions community prediction, which I think Metaculus mostly has. There is already a 7% chance given right now. There are a number of concrete steps that Russia has to or is likely to take before an offensive detonation. If there is a demonstration or ‘test’ detonation by Russia, I would assume the community prediction would jump on the spot to above 20%. Then there is the claim that the logistics involved in getting the weapon ready would take several days and would be observed by American sources, who (based on their track record so far) would likely scream about it in advance. I would also expect a much more explicit threat to come down first. So I’d be inclined to be in the 80% range here.

There is value in helping people calm down and not overreact in these spots, and there is also value in learning how to think about such spots, which is why this is worth our thinking about. I do not think that the practical personal risks yet justify a lot of time spent.

Max Tegmark offers this post putting the chances of nuclear war at 1 in 6 (!) [LW · GW]. I do not buy the logic here, especially his <10% chance that Putin would ‘accept’ defeat (and also the <10% chance that the West would ‘accept’ defeat if it somehow lost). I also think it is much less likely that NATO’s response would be as escalatory as he thinks it would be – my baseline is that the response would be centered around much stronger diplomatic isolation that reduced the need for military strikes inside Russia, also the third step seems overconfident too. On the flip side, he is ignoring other scenarios.

Jeremy Shapiro comes out here and argues that a nuclear war is the expected outcome of the conflict – the whole thing seems misguided, requires Shapiro to be much more insightful than anyone involved in making the real decisions, and is a classic case of not thinking in probabilities let along stepwise probabilities.

Seth Baum here seems better grounded but is assigning no probabilities. Thread contains links to additional good threads.

Timothy Snyder here argues the risk of nuclear use is very low, since it would make Putin’s situation worse. Tyler Cowen found this to be the strongest such argument so far. I agree it is well stated and that its conclusions seem right to me.

Of Boys and Men

Matthew Yglesias reviews Richard Reeves’ book Of Boys and Men. The core observation of the book is that boys educational achievements are falling further and further behind those of girls, which is attributed to a combination of, whether or not among other things, (1) boys being less mature at any given age while school ignores this distinction and (2) a lack of male teachers as a combination of role models and an ability to relate to and make things in school interesting to boys.

Tyler Cowen points with skepticism to finding that gender equality is correlated with boy subjective well-being but not with girl subjective well-being. The obvious response is that the controls on any such study are a true joke, tons of other stuff is going on here, so the result does not at all mean what they think it means. Also, given that in ‘equal’ areas boys are now falling well behind girls in terms of educational attainment, it is interesting that they are also seeing themselves as better off.

These problems are difficult to solve in part because interventions to help kids to get more education have a pattern of working for girls and not working for boys, as Reeves notes in National Affairs. This suggests that as we Do More to get more children more education, boys will ‘fall behind’ more and more.

Solutions proposed by Reeves to boys falling behind are to ‘red shirt’ those ‘not ready’ for kindergarten, who will mostly be boys, and hold them back until they are older, which Matthew interprets as implying universal free pre-K (I don’t think this is obvious at all?), to recruit more male teachers, and to promote vocational training.

We lost more than 20k teachers in September on net, another strong argument that we need to get rid of pointless occupational licensing requirements here.

I see holding children back as another way of saying some combination of four things here:

1. Time-shift boys relative to girls to level the playing field (Reeves’ term).
2. Move towards grouping children smarter than simply doing it by age.
3. Dumb down education further.
4. Keep kids in the educational system until they are even older.

The theory on #1 seems plausible. I’m a big fan of #2, tracking and customization are great, this is a crude way of doing more tracking. I’m against #3 and #4, if anything we keep kids in the system far too long already.

I note that there’s been no mention here or on the podcast I heard Reeves on about whether ‘boys who are more ‘mature’ in the same classroom’ might be worrying to parents of girls for other reasons? That could be a worry especially in high school.

Vocational training is something actual everyone agrees is needed and then no one does anything about it. Presumably for Hansonian reasons.

Recruiting more male teachers is plausibly a pure win other than the politics. The proposal to do this via recruitment and subsidies and favoritism seems wrong to me. The simple solution is to ask why we are seeing so few male teachers, and fix it. That problem is that we require not only college degrees but also educational graduate degrees in order to let people teach grade school, and men aren’t graduating from college let alone willing to sit through what follows these days. By waiving those requirements, we would get better gender balance while also getting rid of a bunch of useless credentialism and addressing our teacher shortage. Win win.

Then there are the other elephants in the room, which are more difficult to talk about these days yet without which the problem is not explained. Why are boys relatively more happy when achievement is higher for girls (and things are generally more equal)? That seems to be a hint that perhaps the reason boys do not Do More in school is in large part that they are choosing not to given what they are offered, and that this choice is making them happier, at least short term.

I would put some of the school-specific parts of it like this:

School is all about negative selection. You get punished for wrong answers far, far more than you are rewarded for right answers. You get punished for disruptions or any deviation from expectations, while you are not rewarded for doing something exceptional. One size fits all. Complete circle. Stay in line. Raise your hand. Ask permission to go to the bathroom. Give the teacher back their password [LW · GW]. Zero tolerance policy.

Grade point average is the score, which is the average of components that are often maximized, which in turn are an average of components many of which are fully maximized, which in turn are composed of components that are basically ‘did you mess this up’ whether it is a homework assignment, a test question, a ‘classroom participation,’ attendance or non-disruption, whatever it is, you name it. You start with 100 points, and then points are subtracted. Almost all of the focus is on what you did wrong. In the extreme, if you have seven classes, six A+ and one F means you get held back, whereas seven C- are fine. And that’s fractal.

Meanwhile, what I would describe as normal boy behavior is pathologized because it interferes with running a classroom, in response to which our society puts huge pressure on parents to drug their kids into a stupor and/or get them constantly high on things they are somewhat pretending are not literal meth.

Men are and always have been higher variance so they do relatively poorly under negative selection.

If you are surprised that boys underperform in this situation, that is an error.

Even more centrally than that, the problem is that we are tying the value of people to their educational attainment.

Yes, women are outpacing men in college graduation. There is unlikely to be anything we would want to and be willing to do that would remedy this, nor is it clear that ‘fixing’ it would matter much for net welfare if the total number of graduates was unchanged.

The problem is that we then take those that did not graduate from college, and we ruthlessly discriminate against them for the rest of their days. Whereas the things that are traditionally valued instead, where men are happy to strive to excel, we no longer value or even punish. If we didn’t do that, this would all mostly be fine, tying back into why we don’t see male teachers even more than we used to not see them.

That still doesn’t begin to tell the full story of what is going ‘wrong with boys and men’ more broadly and I am sure I have several commenters chomping at the bit about all the things I am not talking about here that will point some of them out. I am not discouraging them or others from doing exactly this, it seems like a reasonable time and place to do that. It simply is far beyond the scope of what I am attempting to do here.

I have added this to the list of books I would read if I had lots of time to read books, with my prior on reading it remaining low.

16 comments

Comments sorted by top scores.