Monthly Roundup #14: January 2024

Link post

There’s always lots of stuff going on. The backlog of other roundups keeps growing rather than shrinking. I have also decided to hold back a few things to turn them into their own posts instead.

Bad News

I wonder if it is meaningful that most of the bad news is about technology?

I don’t even know if this is news, but Rutgers finds TikTok amplifies and suppresses content based on whether it aligns with the CCP.


It would be great if we could find a way to ban or stop using TikTok that did not involve something crazy like the Restrict Act. I still think the Restrict Act is worse than nothing, if those are our only choices.

If the CCP limited its interference to explicitly internal Chinese topics, I would understand, but they do not: WSJ investigates the TikTok rabbit hole, in particular with respect to Gaza pro-Hamas content.

Noah Smith: At this point, whether America can bring itself to ban TikTok will determine whether it’s an actual country, or just a country-shaped sandbox for totalitarian states to play in

An analysis of Chinese censorship of American movies. Under their analysis, without such bans we would have 68% of the Chinese market instead of our current 28%. They emphasize factors like occult content, which has an effect but a remarkably small one, only raising an otherwise 50% to be banned movie to a 67% chance to be banned. An R rating similarly takes the odds to 70%, likely largely as a proxy for various things that get you the R rating.

I love buttons that do things. The thing I loved most about early iPhones was that they had a button. A nice, big, physical button, that bailed you out of pretty much anything. Things were simple. Alas.

Matt Palmer: Observation from younger brother: “Whenever I have to adjust the settings on my iPhone I have to Google how to do so, this seems like a red flag.”

Patrick McKenzie: No lie, I had to ask my wife how to turn my iPhone off, now that I have one that doesn’t have a physical home button.

“Isn’t it basically same as it is on an iPhone with a home button?” The thing which stopped was that you need to long press two things but Siri triggers immediately when on button(s) down and I would immediately release them thinking “No I didn’t want Siri.”

And almost every interaction with Settings or any part of the Apple ecosystem is brokered by a Google search leading to Apple dot com or a content farm explaining in four steps which buttons I need to hit. These don’t seem learnable or predictable in most cases.

A decade ago when I started using Macs seriously (quite late in my career for that relative to most geeks’ expectations) I was routinely surprised and delighted by how much the iOS experience on phone/​iPad had prepared me for.

These days iPhone doesn’t prepare me for iPhone.

Can anyone explain why various meeting and calendar apps continuously fail to understand what time zone they are in? I’ve dealt with this a lot as well.

Patrick McKenzie: Why Google’s Calendly won’t crush Calendly’s Calendly in one image. Necessary context: I live in Chicago and am accessing this from a phone which knows it is currently 10:15 AM to schedule an appointment with someone in San Francisco.


Patrick McKenzie: Here are two things Google PMs would say: “The default time zone set in your Google Calendar account is JST. I know a user could have two time zones there, but org politics will not allow me to override the default one.” and “This affects almost no users. Only millions.”

Meanwhile the businesses which actually care about calendaring for power users of calendaring know that many of their favorite users have two, three, or more home time zones and always getting this exactly right is important.

Do they? I am not convinced they do. I am also very convinced that it is utterly insane for a calendar app not to default to the time zone in its current location. It should also be loud about any conflicts, when it sees you moving around or in an unusual location.

Takeovers of phone numbers, especially important phone numbers, are getting worse. The system as it currently exists essentially lets any telecom worker give anyone your phone, and many of them are easy to either dupe or bribe. Meanwhile, everyone increasingly uses phones as account recovery and security, which you have to actively guard against to stop them from doing, and some of them will outright insist.

Twitter Safety: We can confirm that the account @SECGov was compromised and we have completed a preliminary investigation. Based on our investigation, the compromise was not due to any breach of X’s systems, but rather due to an unidentified individual obtaining control over a phone number associated with the @SECGov account through a third party. We can also confirm that the account did not have two-factor authentication enabled at the time the account was compromised. We encourage all users to enable this extra layer of security. More information and tips on how to keep your account secure can be found in our Help Center

SwiftOnSecurity: The attacker uses other channels to enumerate and guess the phone number attached to an account and then checks against the telco they have control over.

The insider only briefly temporarily forwards the victim number to a 3rd party then switches it back to normal once they’re in. This is how they stay quiet since most victims will not have leverage or telemetry to understand how they got hacked. It was their cell phone provider.

Make it so account recovery systems require multiple factors and remove telephony-based recovery for VIP accounts entirely. Go check your systems now. Go try to access all your stuff like you forgot your password.

At a minimum, it is insane at this point to allow verification of anything valuable via only a phone, you need to at least also require another source.

We increasingly care too much about comfort versus other things. But that’s peaked?


From November 2022 (!), 1 in 4 hiring managers said (he admit it!) they’re less likely to move forward with Jewish applicants.

When asked why they are less likely to move forward with Jewish applicants, the top reasons include Jews have too much power and control (38%), claim to be the ‘chosen people’ (38%), and have too much wealth (35%).

Seventeen percent of hiring managers say they have been told to not hire Jewish applicants by company leadership. This is true of more hiring managers in education (30%), entertainment (28%), and business (26%).

And that’s with it improving!

Nine percent of hiring managers say they have a less favorable attitude toward Jews now than 5 years ago, while 31% say they think more favorably of Jews; 60% say their attitude is unchanged.

So yeah, antisemitism was already quite alive and well, all the standard tropes. If anything, that’s still historically pretty good. We have been dealing with this for several millennia. In every generation they try to kill us. We all know Hamas aim at another holocaust. Some people were surprised at who joined the ‘they’ this time, that’s all. I wasn’t.

Sarah Constantin on various reasons she sometimes feels she can’t say various things.

Government Working (USA Edition)

US high skill immigration policy has figured out it can use the O-1A visa for extraordinary ability and also the STEM EB-2 for advanced STEM degrees.

Alec Stapp: Major win for the US on high-skilled immigration policy: “USCIS data show that the number of O-1A visas awarded in the first year of the revised guidance jumped by almost 30% The number of STEM EB-2 visas after a ‘national interest’ waiver shot up by 55%”

If you have an advanced STEM degree and want to put it to work, or have any valuable extraordinary ability, it seems rather insane to not let you come to America and become a citizen. I strongly support doing as much of this as possible.

The rest of the world standardized, but the USA and Canada have their own exclusive standard for elevators, excluding us from global parts markets.

State Farm stops writing new home insurance policies in California due to legal inability to raise prices and massive resulting losses. If you could be stuck selling insurance at or close to current prices indefinitely while facing adverse selection over customers, I don’t see how you can sell insurance priced in a reasonable way.

Federal highway officials hate us, tell local and state officials they must stop using humor and pop culture references on their road safety signs because they might ‘distract.’ That’s the point. You get people to pay attention. Also you brighten up their day. I sincerely despise people who issue rules like this. How do we fight back?

The Farm Bill is mostly subsidized crop insurance. Taxpayers cover 62% of premiums. Which is profitable enough for the farmers that it forces farmers to make decisions that are legible to the insurance, often preventing them from being flexible and adapting to weather conditions or doing proper crop rotations.

This is of course an utterly insane way to do some combination of lowering food prices (which we then try to raise with other programs, and lower again with yet others) and transferring wealth to farmers. It should be up to them how much and what type of insurance to buy. If we want to bribe farmers because we think that’s in our interest to do so or we want to be corrupt, let’s write some checks (or at least give out tax credits) and bribe farmers.

At least it’s not as bad as the part where we also pay people not to plant crops.

Agreed with retiring congressman Patrick McHenry, we need to pay Congress more. I think it was Robin Hanson who I saw say that either you pay them or someone else will pay them, you get to pick which one.

As was inevitable, meet the new Speaker, same as the old Speaker, cutting the same spending deal because of the same conditions, and the same people getting mad about it. Question is what they dare do about it at this point.

California Fatburger manager trims hours, eliminates vacation days and raises menu prices in anticipation of $20/​hour fast food minimum wage. That seems like a best case scenario, unless the goal is to make fast food uncompetitive.

Government Working (UK/​EU Edition)

UK moves to exclude family members from coming in on student visas. The usual suspects pointed out how this is going to discourage students from coming. Nathan Young points out that this is one of those ‘ruining it for everyone’ situations.


The chart clearly shows that this was rapidly transforming into a backdoor immigration mechanism. If the situation is what it was in 2015, something like ‘5% of students take someone along because they need to,’ then you want to allow that. If the ratio starts exceeding 100%, then the policy is being gamed so much it is clearly unsustainable. If you want to allow more immigration, great, but you still do not want to give active preference to those who twist their lives to game the system.

UK’s lawyers advised the government that it was unable to legally discriminate against companies on the basis of their past performance.

Nathan Young: This is disastrous. The UK Government can’t discriminate based on performance. What on earth are we even doing?

Vegard Beyer: Aren’t the rules governing the UK Government’s discrimination between contractors based on past performance… within the sphere of influence of the UK Government…?

Nathan Young: I wouldn’t want to discriminate on past performance so I’m sure they’ll fix it this year.

UK decides what is important to crack down upon.

Emmett Shear: We’re shutting you down. Your pizzas have consistently come in 1/​2” too wide, and we have caught you five times distributing excess pepperonis.


Biggest surprise is that this is a UK pizza photo where the pizza looks edible.

Well, that and any productive activity whatsoever, like renewable energy.

In the past five years, the number of applications to connect to the electricity grid — many of them for solar energy generation and storage — has increased tenfold, with waits of up to 15 years. The underinvestment is restricting the flow of cheap energy from Scottish wind farms to population centers in England and adding to the delays for those with high power needs, like laboratories and factories. Laws that give local planning authorities considerable power are blamed for Britain’s shortage of housing and blocking the construction of pylons needed to carry electricity from offshore wind farms. Residents’ objections to noisy construction and changes to the landscapes have been a stumbling block.

One way the British government turned off investors was by changing planning measures in 2015, and tightening them further in 2018, so that a single objection could upend a planning application — effectively banning onshore wind in England. John Fairlie was a consultant in the wind industry at the time.

Mr. Fairlie is currently a managing director at AWGroup, a land development and renewable energy company that recently got an onshore wind turbine up and running in Bedfordshire, in the east of England, that will generate enough electricity to power 2,500 homes. Because of planning restrictions and grid connection delays, the project took seven years to complete.

It is amazing, and a statement about the expected returns to investment, that such projects still continue at all. Imagine what the UK could accomplish if people were allowed to build houses and generate energy, even if nothing else changed.

Ah, standard plugs.

European Parliment: From 28 December 2024 all mobile phones, tablets and cameras sold in the EU will be equipped with a standard USB Type-C charging port, making it easier for you and better for the environment.


How do they think that works exactly? In twelve months I get rid of all my existing devices? I note all the concerns about ‘what if they had done this five years ago with micro-USB’ and if a new better tech comes along in the future, and yeah, sure, but I’m still inclined to say Worth It at this point.


The map is full of little joys, like Cyprus being in purple.

Trouble in the Suez

It is insane that we are not doing our job of protecting international trade. A bunch of rebels shoot a few missiles, and we can’t stop them? We take weeks to even start responding?

There is a list of things you absolutely do not tolerate as leader of the free world. Disrupting international trade routes is near the top of that list. That’s the job.

Don’t tell me we can’t handle it. Point, counterpoint:

Almutawakkil: I advise Americans and British people to familiarize themselves with some points about the Yemeni fighters ( Houthis) before rushing into anything.

– They don’t follow your movies and TV shows at all.

– They are not bothered by your media or social media distractions.

– Psychological warfare is utterly useless against them.

– They are natural-born fighters, really, no kidding.

– Their life goal since childhood has been to fight America.

– The last will and testament passed down from their ancestors is to liberate Palestine.

– At the very least, they have 4 to 5 wars of military experience in various terrains.

– They have all written and recorded their life wills in both audio and video formats.

– The martyrdom of one of them is a tremendous source of pride for their children, family, village, province, and country.

– Their poets passionately glorify war more than any love, flirtation, or romance poetry.

– They all obey their leader, Abdul-Malik Badr al-Din al-Houthi, with absolute obedience.

– Their only fear is the punishment and wrath of Allah if they fail to support the people of Palestine and backtrack on their support.

– They love death as much as you love life, if not more.

In any confrontation they engage in… I won’t explain these words… you will come to know, understand, and feel them more when facing them.


Frank Fleming: I advise foreign countries to familiarize themselves with some points about United States citizens (Americans) before rushing into anything.

– They enjoy multiple streaming services.

– Each day they get worked up and outraged by something on social media that would be impossible to explain to you.

– Psychological warfare works really well on them but only for a few seconds before they get distracted by something else.

– They probably have no idea where your country is and maybe have never even heard of it.

– Their life goal since childhood is to be a popular influencer.

– The last will and testament passed down from their ancestors is to prefect their BBQ recipe.

– At the very least, they can walk up two flights of stairs before being winded. – They have 401ks, but probably not enough in them.

– Getting a post to go viral is an extensive source of pride among their community.

– Their poets passionately glorify getting superpowers and fighting supervillains.

– They only elect the dumbest idiots as leaders and never listen to them.

– Their only fear is their phone running out of power when they’re away from home.

– The only reason their enemies are still around is it feels unChristian to completely obliterate them.

In any confrontation they engage in… I won’t explain these words… you will come to know, understand, and feel lucky if your entire effort to fight against them merits you to even be a future question on Jeopardy!.

History is littered with tribes who studied nothing but war wiped about effeminate guys in white wigs. If you want to defeat America, learn to code or something.

There was a time, for thousands of years, when ‘we do nothing but fight for generations’ was the way to go to win wars. When the dudes on horseback periodically sacked the cities and became the new ruling class. When it was said, as in the end of Herodotus, let us live somewhere hard so we might win wars.

Now, not so much. I may not be ‘appreciating the complexities’ but if I am Biden I get on the phone, explain that either shipping is going to resume or there are not going to be any more rebels, as an example to the next ten generations, and I mean what I am saying.

We did not go that far. We did eventually start using force.

Bret Devereaux: There is a sort of performative naivete for the folks acting shocked, shocked! that it turns out that disrupting more than 10% of all global trade does, in fact, lead to a kinetic military response. Of course it did.

And just a reminder for the folks who think this is about Israel – the Houthis have been firing on ships indiscriminately. If they were just attacking ships bound to or from Israel, I doubt we’d see the same level of response.

You do not get to pirate ships chartered by Japanese companies to move from Turkey to India because you are mad about Israel. You don’t get to try to seize Danish ships moving from Singapore to Egypt because you are mad about Israel.

Or, well, you can, but then this happens.

I continue to be surprised and dismayed that we have not done more. The situation is completely unacceptable. Anyone who has an issue with using force to stop pirates, or thinks that the actions of unrelated nations could possibly excuse it whatever you think of those actions, can go to Davey Jones’s Locker.

It does seem that on the 22nd we did another set of airstrikes. This still does not seem to appreciate the stakes:

Jim Bianco: ~70% of all shipping is conducted on a long-term contract. A cargo ship is essentially a shuttle between ports. If they have to go around Africa, that adds 20+ days to the route.

So, if a ship can make six runs yearly, the extra distance means it can only do four or five runs yearly under current conditions.

To make up for this shortfall of runs, excess shipping capacity is contracted on the “spot” market. This chart shows worldwide “spot” rates are up 85% in the last two weeks, the largest two-week jump (bottom panel) since Drewey started its index in 2011.

Shippers are aggressively grabbing excess shipping capacity and will pay up big to do it.

The objective of the military action against the Houthis is to allow unarmed commercial ships to sail the Red Sea with affordable “war insurance” rates. These rates are up 300% to 500%.

I FEAR we are weeks or months away from commercial shipping returning to normal in the Red Sea. Until then, supply chains remain snarled, and the inflation pressure on goods is very real.

Meanwhile, the propaganda wars got weird. Why are we having propaganda wars where one side are literal pirates? How is this a call people are in doubt about?

Daniel Eth: Describe the last 500 years of great power conflict in a tweet:

Kane: the funniest part of the red sea houthi pirate conflict is that the pirates keep posting super macho propaganda videos only to be annihilated while the captain of the carrier doing the annihilating is just tweeting about cute dogs and stuff

Chowdah Hill: This captain only loves me for the snax. I was hoping for a more productive working relationship, perhaps a few team ups or something. Instead… just snax.

There are those here who are cheering on the rebels for trying to disrupt shipping. These people are enemies of civilization and of humanity. Treat them accordingly.

Crime and Punishment

Periodic reminder: The rate of rape in prison is almost 5% per year, the majority of sexual abuse reports were by staff rather than other prisoners. It is pretty stunning that we all continue to accept this as part of our justice system.

If someone is indeed saying this (the video won’t load), many things have gone very wrong.

EndWokeness: Canadian police warn residents not to post photos of thugs stealing packages.

“You cannot post the images… we have a presumption of innocence & posting that could be a violation of private life” -Comms Officer Lt. Benoit Richard

If the police are unwilling to do their jobs and arrest people who steal, as often the police are unwilling to bother to do, the least they can do is not actively get in the way. You have a presumption of innocence in court, and only in court. Even if that was not true, a presumption of innocence does not mean no one can accuse you, and no one can post evidence. That is completely absurd. As is any ‘expectation of privacy’ while stealing a package off someone else’s private property.

Poor people commit more crimes. Alex Tabarrok asks, why? He points to a Swedish study by Cesarini et al, studying lottery winners there. Winning the Swedish lottery does not substantially decrease crime despite it paying out over time and looking a lot like a permanent income shock. This continues the pattern of lottery winners proving largely unable to use their money to get better life outcomes. I do not think it translates zero to other questions, but lottery winnings being very clearly luck and happening all at once I do think makes them categorically different.

The cost of crime is high, even when it does not happen to you.

Audrey (of San Francisco): I used to take dance classes at a studio on Market st 3-5 times a week. I was perplexed by people who would pay $25 for a 50min yoga class when a 90min ballet or jazz class with live music cost $9. I would usually jog there and walk back, but then the area got more sketchy so I started to call Ubers there (which made the yoga classes comparable in price).

Then the area got SO scary I basically go 0-1x a week (to just one ballet class during weekend day time since the instructor is dear to me). Meanwhile the building put metal over their glass doors and now has at least two people to guard the door and manager elevator.

I can’t imagine how hard this is for the dance studio to need to spend more money for building security and have fewer dancers come. I am also begrudgingly taking more yoga classes that are boring and expensive because I can walk there and back without having to dodge needles and people on some horrible drug shrieking and violently flailing around.

Nix: I think I know this dance studio… stopped going for same reason. The side street was so rough especially as it got dark (like people screaming etc)

I pay a huge portion of my discretionary income so my family can live in New York City. If crime was the way it was when I was growing up, my willingness to pay that would go way, way down. Luckily, things are much better.

Illinois eliminates cash bail. It seems the plan is to not charge bail, hope everyone shows up anyway and that it will all work out?

George Washington University law professor Kate Weisburd said in other states that have implemented bail reforms, like California and Texas, the use of ankle monitors has gone up while jail populations decreased. She said an increased reliance on monitoring isn’t “moving the ball forward when it comes to pretrial justice.”

“I think what makes the [Illinois law] so powerful is that judges are required to release people who are deemed not to be a safety risk and not likely to flee,” Weisburd said. “So that means that most people released under this new law don’t need to have an electronic monitor, because they’re not a safety risk, and they’re not a flight risk.”

I notice I am confused. How is going from ‘put you in jail’ to ‘have you wear an ankle bracelet’ not ‘moving the ball forward?’ That seems like moving the ball forward to me. Wearing an ankle bracelet is at least an order of magnitude less bad than being held in jail? I would say at least two? And for many people, far better than paying the bond to post bail even if they could? I mean, you could pay me to wear an ankle bracelet and it would not even be that expensive.

As always, people confuse ‘not available’ with ‘not available at this price’:

Garrison said even if they had more money, there aren’t attorneys available to hire. Macoupin County is part of Illinois’4th Judicial District. It includes 41 counties in central Illinois. This year, only 55 new lawyers were sworn in in the 4th District, fewer than 112 attorneys per county.

There are tons of lawyers, by all accounts, who are in need of work. AI will likely streamline much legal work further, expanding that pool. Do these people want to go to Macoupin Country to work with criminal defendants? No, mostly they do not want to do that. Also, if you raise your price, some of them will do it anyway.

RCTs on interventions in criminal justice almost always show no benefit. The obvious follow-up is, suppose we did anti-interventions, would we expect to see no harm?

What happened when judges were given algorithmic risk assessments on defendants, while still having discretion to make final decisions on sentencing?

Megan Stevenson (paper author): We find that the judges DO use the risk assessment tools, but mostly only during the first couple of years after adoption. After that, they seem to stop consulting them.

But even in high-use periods, they overrode the recommendations associated with the risk assessment frequently!

Although the risk assessment was implemented solely for the purpose of diverting people from prison, it had no effect on incarceration rates.

There are some curious expectations at play here. Megan seems surprised that judges frequently ‘overrode’ the recommendations, despite the recommendations being based on only a subset of the factors judges care about and considering only some of the evidence, and also judges being humans who think they know better.

Megan also seems surprised overall sentences stayed the same. Whereas of course judges are not going to think risk assessments should alter how tough they are on crime. Good job judges making the proper calibration adjustments. Yes, if you say some people are low risk hoping those people go to jail less, the ones it says are high risk will then be put in jail more.

Megan Stevenson: Below, we compare the *actual* impact of risk assessment in the hands of humans to the *simulated* impact of sentencing by risk assessment alone (no discretion). [shows graph with no impact on average length of sentence]

Deviation from the recommendations of the algorithm is systematic: longer sentences for Black defendants and shorter sentences for young defendants.

Risk assessment had not impact on racial disparities, likely because judges already sentenced in a racially disparate manner. It led to harsher punishment for young defendants — but human discretion mitigated the full negative impacts on young people!


I read this as: Judges care about things your risk assessment does not. They think younger people, and women, deserve consideration, for reasons that are not about risk.

Not sure what the story is here regarding unemployed? If I had to guess, the judges noticed (consciously and systematically, or otherwise) that the risk assessments made unemployed people very high risk, and did not think that was equitable or something they should get punished for so much, so they scaled it back.

What about black defendants? Certainly there is some amount of racism involved. There is also the possibility that the risk assessments deliberately ignored or controlled for various factors to correct for racial disparities or ensure equities, and the judges learned to correct for this or simply observed the facts and overruled.

Stevenson is framing this as ‘we had a risk score, and they overruled it.’ I am confident the judges instead were thinking ‘ah, good, a risk score, we can try using this as one of our considerations.’

If you thought this could convince a system to stop being racist, or stop putting people in prison so often, I would wonder why one would expect that to stick?

Instead, the risk scores worked in doing the thing one would hope, which is moving incarceration from those with low risk scores to those with high risk scores.

In sum, risk assessment use in the hands of humans led to a reshuffling of prison beds — no net decline, but a shift towards incarcerating those with higher risk scores and releasing those with lower scores.

And yet, this didn’t work?

Theoretically, this should have led to lower recidivism rates, since the highest risk people were locked up. This did not happen. We can reject even small declines in recidivism.

So what is going on there?

Why not? Maybe the tool had less novel information than expected. Maybe judge’s used it in the “wrong” way, over-riding it when they shouldn’t.

The tool meant more emphasis on the factors considered by the tool, excepting those undone intentionally by the judges, and less emphasis on other factors. Yet this did not help.

I find the ‘over-riding it when they shouldn’t’ hypothesis unconvincing. The model predicts that things should have improved given these choices. Things did not improve. Judges would have to be doing far worse than random, in terms of recidivism, in deciding when to overrule.

But “wrong” is subjective. If the only goal is preventing crime via incapacitation, teenagers should get the longest sentences. Young people are by FAR at the highest statistical risk of crime.

But there are lots of goals at sentencing. And many people — Virginia judges included — don’t love the idea of harsh punishment for teenagers.

In Virginia, discretion mitigated some of the adverse effects of risk assessment (harsh sentences for the young) at the expense of its benefits (reduced incarceration/​recidivism).

Quite so. This is certainly a reason to expect judge final decisions to score worse than the algorithm on risk alone. But it would still predict that, given you saw a shift in who got sentenced from low risk to high risk, an improvement in results.

So the algorithm has some explaining to do. Why were judges unable to improve the production possibilities frontier?

Why did the judges ultimately decide the scores were not useful? Notice that they were correct about this.

To be useful, a risk score has to tell the judge something they do not already know. So we’d need to look at what makes up the scores. What is the new information?

Good News, Everyone

Adam Grant suggests: “I’m giving you these comments because I have very high expectations for you, and I’m confident you can reach them. I’m trying to coach you. I’m trying to help you.” Then you give them the feedback. Love it.

Wind turbines are friendlier to birds than oil and gas drilling, purely in terms of directly damaging wildlife. And of course they are four (yes, 4) orders of magnitude less deadly than cats. A sane civilization would have a blanket ‘no you do not get to say what about the birds’ rule in place, certainly not if the particular bird is not endangered.

Claim that solar power and energy storage will eat all other power sources and reach total dominance. Certainly if you continue on an exponential for long enough that is what will happen. Predicting total dominance is a much better prediction than the continuous official predictions of linear increase every year.

Cost per hour for various digital media. Essentially all TV and video subscriptions are bargains for the average user, as is Twitter. The only issue is that this makes us unwilling to pay for the movies and shows we actually want if they’re not included, I am learning to stop doing that but it is tough. Games he treats strangely, with $60/​game and also assuming very long play times. Games are reliably a bargain if you like them, the trick is finding the right games for you. That’s true for basically everything here. The real cost is always your time.

The Puritans would one-box in Newcomb’s Problem. So what if the decision on whether you are Elect has already been made and what you do now can’t change that? Have a good enough decision theory to do your best anyway. Generalize this!

Suhail notes a curious effect.

Suhail: One thing I’ve noticed that drastically reduces my screen time is not allowing my phone to be in the same room as I sleep. Unsurprisingly it’s the first thing I’ll reach for and I’ll clock in 30-45 min. What’s been surprising is how much less I’ll reach for it throughout the day.

The other benefit is it protects my mind. It’s subtle but if I read a tweet, news article, etc I’ll start thinking about it. If I wake up to my own thoughts, I find those thoughts far more satisfying to begin the day. Maybe it’s a personal thing or a project I was working on.

Everyone gives the ‘don’t have your phone there’ advice and almost no one follows it. I do believe I have gotten pretty good at not actually using the phone while it is there without a good reason, but there is a clear effect where doing that still requires effort. The part that is interesting is that he reports this also helping throughout the day.

Note that ‘out of the room’ need not be literal. Technically my computer and work area are within the bedroom. Leaving the phone there would be distant enough for me.

Emmett Shear threads on agency and how to cultivate or teach it. A key suggestion is ‘write down the dumbest plan that could possibly work’ to avoid having to find a plan that will work, and still verifying that your efforts could, somehow, end up working. Other good questions include ‘what’s the stupidest easiest one thing you could do to make even a little progress?’ ‘What if it was possible? What might be a good first step?’ and ‘It sounds like you’re sure you won’t succeed, what’s going on with that?’

He says agency is a complex skill. In some ways it is. In other ways it is simple. Or, it is functionally complex, but conceptually simple.

Modern elevators have overlapping failsafes. If the cable snaps, then most of the brakes would have to fail, and even then compression of air and the springs at the bottom should mostly prevent injury from a freefall.

JOMO, the Washington Post says, is the Joy of Missing Out, and you should cultivate it more. I was ready for a historically bad take. Then I got a good one, which is that ‘missing out’ on social media in particular is good, go live your life. You want to fear missing out on real activities, especially in person. You want the joy of not looking at your phone.

Bernie Sanders again quoting the claim “63% of Americans do not have $500 in the bank to pay for an emergency healthcare bill.” The good news is that this is obviously false. Median household net worth is $192k including $8k in checking.

Rampant corruption in Chinese military procurement led to purge of army, Bloomberg says, with missiles filled with water instead of fuel.

NPR reporter fired for ‘offensive’ stand-up jokes, was forcibly rehired because arbiter decided jokes were funny.

This seems true, and I have occasionally done this:

Paul Graham: A lot of essay writing is not so much telling people new things as helping them to reach conclusions they were already 90% of the way to themselves. It’s easy for an uncharitable reader to dismiss such essays as obvious.

That’s 90% true. And yet false; that last 10% is hard.

Sports Go Sports

Nate Silver is optimistic about the new Las Vegas A’s.

I strongly agree with Tyler Cowen and his reasons that we want to keep sports teams playing within city centers. You want to encourage people to make trips to the city center. You want to enable people to combine trips to multiple locations. You want to allow easy transitions in and out of the stadium. You do not want to be locked into only the team’s offerings.

Location, location, location. All of this is vastly more important than a nominally nicer venue. I love Citi Field. It is an amazing ballpark. I would still happily prefer a lousy ballpark that was closer and within the heart of the city. And I would happily take the old lousy Shea Stadium over a Citi Field (or even the platonic ideal of a stadium) if the new place was not on a Subway line, or on a much less accessible subway line.

NBA in-season tournament is a big hit, everyone loves it. I agree that this is a great development and we need to see more things like this. If they never flop, we are not running enough experiments. What sports needs are storylines, stakes and motivation. With the expanded playoffs in every sport, if you don’t do anything to fix it, the regular season loses meaning. The NBA should also flat out reduce how many games they play, but there are understandable reasons they don’t.

NFL players go bankrupt at a constant rate regardless of how much money they earned over how many years. That is super weird to me. The amount of money really should matter, yet somehow it doesn’t? It is really hard to be that bad with money.

ESPN used fake names to get unearned Emmys for many of its stars, including those on College Gameday. It seems like what they actually did was get them Emmy-shaped physical statues which they never earned? Which is hilarious, also who cares. There is a very clear record of who did and did not earn one. An unearned trophy is nothing.

Ben Krauss calls for reform of sports betting, saying that the combination of mobile betting, aggressive notifications and other advertising tricks is increasingly causing big problems. It is a difficult balance to strike, but I agree things need to change. I actively like that College GameDay discusses point spreads and has someone making a few picks. I do not think it is fine that people are getting lots of in-game push notifications. Charles Barkley should not be able to, on television, offer ‘guaranteed parlays.’ Letting people bet on their phones is clearly dangerous at best. The balance is tricky.

One place the industry continuously offends me, that does not offend Ben Krauss as a purely casual gambler, is the prices. With the epic growth in gambling volumes, and the ability to bet in person with low transaction costs, we need to see a lot more competition on price. Alas, regulatory and advertising costs, and the cost of deposits and withdraws, are standing in the way. It is still insane and kind of criminal that ESPN is showing us truly obnoxious baseball lines that go −120/​+100 or worse as if that is an acceptable thing to do.

As Seth Burn put it, math is not this hard.

Kirk Herbstreit: “I think the 12-team playoff is going to create a lot of buzz,” Herbstreit said on College GameDay. “How many games will that be, seven total?

“I think you eliminate the bowls,” Herbstreit added. “Nobody wants to play in them, don’t play the bowls. Just have the 12 teams—we’ll get excited about those—and if you want to add maybe five or six bowls outside of that, then do five or six. But we’re getting to a point where it’s ridiculous.

Kirk is actually pretty great both on GameDay and as one of the best full-spectrum play-by-play announcers. I agree that there are far too many bowls. You should only get a bowl if you accomplish something, which does not mean going 6-6. I think it would be fine to say you need either 8 wins, a conference title game or the top 25?

Tony Hawk one year made four million dollars off the Tony Hawk Pro Skater games.

Gamers Gonna Game Game Game Game Game

Magic: The Gathering bans some cards. Channel Fireball’s LSV reacts. It is odd to read about such developments while this removed from the game.

Magic: The Gathering Arena introduces Timeless, their version of Vintage complete with original versions of all tabletop cards and an actual three-cards-only restricted list of Channel, Demonic Tutor and Tibalt’s Trickery.

Brilliant, passionate and scarily accurate thread from Cedric Phillips about what drives Magic players to attend tournaments. Decklists, feature matches, deck techs, chance to make your name, narratives and excitement, aspirational experiences, staying at top of the circuit. Not the prize money. Amazing points. Also someone hire this man please? He is very good at this sort of thing. Alas, I have nothing relevant for him to do.

I am not as down as he is on the importance of prize money, you need to give them that kind of hope too, especially if you want to let people turn fully pro. You also need enough to drive the proper attention and prestige, so they feel real. But what matters to people most is attention and prestige. Ben Seck confirms. Brian Kowal confirms. Sam Black confirms, was was never focused on asking for more money, but as he noted he made his money off content creation. LSV confirms that switching from aspirational to esports and entertainment was deadly, players need to think that could be them.

I continue to think Magic would get a huge ROI from a true return to form of the Pro Tour including very large prize pools. But to make it work, all the prestige stuff has to get knocked out of the park too.

Selling slots on a Magic Pro testing team for $300 is either way too much or way too little. The amount of labor and value here is intense. You’ll spend a lot of time with at least one dedicated pro. So either this is a sacred value that must be $0, or it is worth way more. I lean towards the latter. There was basically never a point at which I would have let someone I didn’t otherwise want onto my team this cheap, and I’d happily pay $300 for someone else to be handling all the logistics.

Crypto trader withdraws $25 million worth of ETH by spending it all on Magic: the Gathering cards that got handed to him in person. Patrick McKenzie is both offended as a geek and respects the genius of the move, where you buy an object you can move physically, using payments that look like product purchases, that then trades like a gold bar, without screaming ‘I am a gold bar.’


Advice to anyone building a new rogue deckbuilder is to not make it easy to assemble tiny decks, or to do something to seriously punish anyone who does it.

Jorbs reascends the Spire from scratch, going 80-3 on ascending over about 80 hours, with 3 additional losses in act 4 for 70-6 (since the first three runs weren’t allowed access to Act 4). One of the losses outside of act four (A17 Watcher) sounds clearly avoidable if not goofing around, the other two sound like whammies. He notes biggest difficulty spike was losing third potion slot, other notables are Ascender’s Bane, gold hits and worse events. He didn’t much notice stronger enemies, whereas I do notice, he notes that is likely a reflection of how he builds. He also notes he had fun playing janky decks that don’t work on A20. As he noticed right in his first run, the problem with such runs is that you spend a lot of time going through motions of runs you’ve already won, which is also the issue with many daily climbs.

Interview with Jonathan Rodgers, co-founder of Grinding Gear Games, about Path of Exile 2. He says that loot can only have value if it might have value to someone else, hence you must enable trade. I thought Diablo 3’s auction house proved the opposite, that if you allow trade then loot only has value that it holds in the marketplace, which means loot mostly has no value. The variance disappears, you can always trade for items that get the job done. Whereas if you are looting for yourself (e.g. Solo Self-Find, or at most a small group) and there is no fungibility, loot becomes more interesting.

I strongly agree with him to stop with the +2% modifiers, +20% or GTFO, you want to make sure everything each item does counts and you can feel it. I also agree on the power and necessity of the reset button, to strongly encourage everyone to start over.

I’m very much looking forward to Path of Exile 2. Path of Exile is far and away the best Action RPG of all time, and the only one I’d put in my Tier 1 of Must Play (I’d have considered putting Diablo 2 there, if Path of Exile didn’t exist, but it does.)

Exodus sounds like it’s going to have some cool things to do with time dilation.

Emmett Shear reminds us that if you are playing Street Fighter [2 Turbo, presumably] then the solution to the so-called ‘cheese’ moves that seem overpowered is not to ban them, it is to use them until someone shows you or figures out the counter, then everything is fine.

This works exactly because the game is well-designed, with good counters to every such move. If that was not true, this would fail. It also relies on having enough data to find the counter-moves, and enough practice to learn them, to get to the new equilibrium. It does genuinely ruin a different experience some people want. Keep those things in mind while generalizing.

China announces planned restrictions on video game monetization. They intend to ban daily log-in rewards, bonuses for first-time spenders, incentives for repeat 5spenders, not having a spending cap, offering loot boxes to minors, not letting items be purchased directly, and the auctioning off of game assets. Also unspent currency must be refunded at purchase price if a game shuts down.

Bravo. Mostly. I notice that there is a problem with Magic: The Gathering and other tradable or collectable card games. It would be nice to find a way to exempt sufficiently ‘real’ games. I presume Magic: The Gathering Arena and Modo can survive this in China, but it will be tricky. Emergents, had it survived, would have had to either leave China or radically change its economic system.

That is still a price I would be willing to pay. Gacha (I will always call this Gotcha in my head) and gambling games, and dopamine-based tricks like daily logins, are the bad money that drives out good due to how mobile customer acquisition works. Despite all the obvious reasons to be opposed, I think this is sufficiently good for human flourishing that I am fine with it.

Game Reviews

Mahokenshi was a fun little game. I did a relaxed pace, no-information full-achievement run in about 15 hours. Think rogue deckbuilder, with a very small deck, on a hex grid with goodies and enemies, usually against a clock. I rank it Tier 3, worthwhile for fans of the genre, with two caveats. The first is that the game is not difficult. The other is that there is a huge lack of balance between the four characters or Samurai houses. One is very obviously busted, especially going for many challenges where you need to go fast. Then again, if you want the game to be more challenging, one way to do that is to say you have to rotate between the houses you can play, and then you can’t use the broken house (you’ll know which it is) once all four houses are unlocked.

Cobalt Core is a fun little roguelike deckbuilder in small doses, and it has its charm, but ultimately I can only put it at Tier 4. There is not enough variety in cards, strategies or enemies, you often know you’ve won a run before the first boss, there are severe balance issues and the game doesn’t encourage you to do challenging things, with the highest level being more ‘you randomly die easily’ than anything else and the game not gating anything behind playing on it. And it asks you to play way more games to unlock things than is reasonable. With some more work this could be Tier 3, but in its current state, diehards only. But did I have some fun? Sure.

I played a bunch of Backpack Hero. I wanted to like this game a lot, but ultimately can only classify it as Tier 4, for diehards only. I had fun with the core concepts. Alas, the balance was all off. It took quite a long time before I was in any danger of dying. When I occasionally did, it felt like carelessness, until I moved to secondary characters that had it much harder, were far more fiddly, and that I enjoyed less. You had to do a lot of runs before things unlocked properly. The powerful things are stupidly powerful, many options seem highly under-developed. The first two heroes are straightforward and fun at their core, the next two felt fiddly and not fun.

Octopath Traveler II is my current game, so I don’t yet know if they stick the landing (I’m wrapping up the first few of the individual stories now with the main party around level 51), although other reviews hint that it does. The first game didn’t lay sufficient groundwork for the real ending, whereas I am pretty sure I know more or less where the second one is going. Did you like Octopath Traveler? This is more of it, seems to be improved around many margins. There are a few places where one could reasonably say ‘are we really doing this again?’ and yes you are doing it again but that is mostly fine. It is impressive how the whole is greater than the sum of its parts. The flipping between stories makes them work. You do have to be in for a long journey. My guess is this is on pace to be Tier 3 but fighting for Tier 2.

I Was Promised Flying Self-Driving Cars

Waymo crash data shows only three injuries in seven million miles, all minor, much lower rates than you expect with human drivers. They only generate 25% as many insurance claims as human drivers and generated zero injury claims. This does not tell us much yet about fatal crashes since those are one every 100 million miles, and tail risk could be different if there are weird failure modes, so the question is whether there are rare weird failure modes.

While I Cannot Condone This

Not enough links? Astral Codex Ten’s monthly links are here, only a few are things I’ve linked to here or otherwise.

Americans do not read many books. Even listened to counts here.


It makes sense to me that not many people read exactly one book in a year. Once you’ve read one, about half the time you’ll read more than six, and half of that time you’ll read more than fifteen.

A fun study found via MR of how long chocolates last in hospitals. This is one case where it should have reversed its final statement and said ‘further study is not needed.’

I had a whole Christopher Alexander sequence planned before AI happened. There’s so much good stuff there, I am still glad I read A Pattern Language.

Made in Cosmos: Christopher Alexander is so wild. 80% of his ideas about home design make me go “wow, how come I never thought about it before?”, and then he’ll randomly come up with something like putting guest alcoves in your master bedroom so that you can all have big sleepovers together.

Charlie Page: Are you saying that’s not a phenomenal idea? Master bedrooms are too big anyway.

Made in Cosmos: lol our entire apartment is probably the size of an average American master bedroom. I dream of a time we’ll have a bed that can be approached from both sides.

A Pattern Language is very clear that not every pattern fits into every house. You choose the patterns that have the most value to you, that fit your space and your life. Also yes, alcoves in the master bedroom are an awesome idea if you have a lot on which you can build a non-standard structure, and therefore can choose to add alcoves. Remarkably efficient use of space to generate optionality. As Cosmos notes, not applicable for everyone, but also it would be a very good way to get extra beds into a tiny footprint if that was your puzzle.

Is this the year?

Paul Graham: Prediction: Wokeness will recede significantly in 2024. There were always more people against it than there seemed, but many were afraid to say so. Now that it’s safer to criticize it, more will.

Manifold traders say 39%, which is pretty good for a substantial move in one direction.

I mostly tried this for a few years. In my job it didn’t take.

Paul Graham: I don’t think journalists or universities grasp how much their reputation has suffered, and that it’s due to their own intellectual dishonesty. A generation ago newspapers and universities were esteemed institutions. Now you see open contempt for them.

A journalist seeing Suhail’s tweet would presumably think “nutjobs are always saying things like that.” But Suhail is not a clueless extremist. Exactly the opposite. And yet is there any journalist in the world who can even see, let alone admit, that there’s a problem?

I mean, not quite the opposite. I’ve seen his views on AI. He does, from what I can tell, support building smarter than human intelligence as quickly as possible and letting it proliferate and thinks that would be good for us. He quoted his company’s written testimony to the House of Lords with pride, in which they commit outright fraud regarding the ‘integrity’ of their investment portfolio’s AI products, claiming we now understand such AI models. But definitely not a nutjob.

Amjad Masad (CEO Replit): Agreed, but what’s the alternative to find ground truth? I hoped Twitter/​X + Community Notes + Free Speech + Transparency would be it. But it’s neither free nor transparent, and notes are easily gamed.

Paul Graham: One way is to follow people whose judgement you trust.

Andrej Dabrowski: It doesn’t scale though.

I disagree, Andrej. I think it scales fine. If everyone has a pool of people they trust, but is doing the work to adjust that pool to get it right, that absolutely scales. In my model, everyone has a ‘level’ (from 1-4 or so) of sense making production, and your goal is to follow people one level above you and those at your level, make sense of the worthy ones, and then make sense to those at or below your level in return.

Journalists used to be accepted into this as All-Level sources, without much question, in a way that rewarded reliability and allowed everyone to understand. Now they’ve lost the necessary faith in that institution. You need higher-level people you trust to be able to use Bounded Distrust on the outputs. Thank you for putting some of that trust in me, keep an eye and ensure I stay worthy of it.

Andrew Gelman reweighs himself on his bathroom scale 46 times to compute the standard error. I mention this partly because it is inherently cool, and partly to tell the story that you cannot do this on my bathroom scale. If you do, you will get an answer of zero. It will come back the same every time.

Is that because the scale is super accurate, or at worst off by a fixed amount? Oh, no. Nothing like that.

It is because someone decided that the scale should have memory. If it gives you 161.3, then it has decided that everything from about 160.9 to 161.7 is going to count as 161.3 for a while. You can even see it, sometimes, bouncing towards the ‘real’ number, then at the last moment it reverts to its baseline. So if you (for example) were to pick up something weighing 0.2 pounds before weighting yourself, then weigh yourself again without it, you’d get an answer 0.2 pounds higher than otherwise.

I am fascinated by who thought this was a desired behavior. Writing this inspired me to get a second scale, for now keeping both around because it is fascinating.

You want to complain? I want to complain about all your complaining. Or do I?

Owen Cyclops: There’s a culture divide you can go your whole life without pinpointing: groups where complaining is negative, and groups where complaining is a normal positive method of socializing. they cant understand each other. larger than a language gap. probably best if they never interact.

Emmet Shear: Games People Play names a bunch of these games, like “Wooden Leg” and “Ain’t It Awful.”

Lilibeth: I’ve found that the ones who don’t tend to thrive in the cultures of the ones who do. Mainly because they don’t know how good they have it, and so the ones who don’t can lap up all the good things. And thrive.

Ben Linzel: Those groups are called men and women and civilization is built around pairing them.

Emily: Been thinking about this all day with shame about my whole family’s complaining culture. So far I have not complained today and I’m going to try actively not to anymore. This tweet bodied me with embarrassment.

I would divide complainers into two key subcategories. One we could call the commiserators (or simply the complainers, or if you want to treat them with proper disdain rather than be even-handed, the whiners), the other the critiquers or the optimizers. The first group wants your social attention on the complaints they are making, the second group wants to fix the problem.

Then you can also divide the non-complainers. You have those who do not complain because they are in Guess Culture, and you have those who don’t complain because they choose to instead not expect their complaints to be heard, at least at this time. They don’t expect you to figure it out or tell you implicitly, they don’t ‘drop hints,’ they suck it up, do what needs to be done and keep things positive. The first group wants your attention on their complaints they aren’t making, the second group does not.

I love the culture where it is standard to critique and complain about everything in a good natured way. Magic: The Gathering culture is like that. When I was gambling it was like that. Rationalist culture is often like that.

Over time, I have also grown to appreciate the need, often, to prioritize a nice time and keeping things positive. You still need to strike a balance in a way that often doesn’t happen, where when it is sufficiently important you speak up. But yes, there is something pretty great about there being times and places to sit back and enjoy, and not be optimizing or complaining and not getting nerd sniped by everything.

There is also a time and place to enjoy a good rant, and loudly complain about how awful things are even if you don’t have a larger goal in mind. In small well-timed doses this is great. When people make it a habit or can’t stop or take it too seriously? Not so much.

There are also times when one must stop complaining because the social punishment would be too large, and find ways to indicate your information and preferences when you can. I hate this. The ‘upper classes’ seem to largely operate this way in most times and places, playing these comedies of manners, and I think this alone is bad enough that you mostly shouldn’t envy them. Their lives seem rather worse than mine.

Money Stuff

I mean, I love it, too perfect, so even thought you’ve all seen it by now:

Gary Gensler (January 9): The @SECGov twitter account was compromised, and an unauthorized tweet was posted. The SEC has not approved the listing and trading of spot bitcoin exchange-traded products.

The ETFs were ultimately approved.

Vitalik Buterin offers financial advice, much of which many in crypto need to hear:

Vitalik Buterin: [not diversifying] is awful advice. Some actual financial advice:

* Diversification is good.

* Save. Get to the point where you have enough to cover multiple years of expenses. Financial safety is freedom.

* Be boring with most of your portfolio.

* Don’t use >2x leverage. Just don’t.

Nothing I ever say is investing advice, but I agree, especially about the leverage. I would add a general principle that one should not worry much about the details of things like diversification or ‘balancing.’ The point, once you have enough savings that it maters, is not to die on any one hill even if that hill is Nvidia. Or if that hill is cryptocurrency. I do not care how bullish you are, there is no reason to risk ruin.

We were promised a recession. Tyler Cowen reminds us of this, asks why we were promised one that then never arrived. As he notes, the correct response is to notice the confusion, not to sweep it under the rug or pretend you made a better prediction. Scott Sumner notes that this seems to be due to aggregate demand stubbornly refusing to fall. I did not predict a recession, but only because I did not make a prediction at all. No points.

My hypothesis is a little out there, and of course Cowen’s Third Law that all propositions about real interest rates being wrong applies, but my hypothesis is that this is not unrelated to AI.

Everyone keeps saying that expectations for AI should raise real interest rates. Well, what if they did raise real interest rates? Not a ton yet, but some. The mechanism is for now only a little bit productivity and consumption effects, although we do have a few areas like coding. It is mostly investment and the anticipation of future investment and opportunity and growth, leading to consumption smoothing and also greater willingness to borrow and such, and people who place bets on future rates impacting rates now. Real monetary policy is not a number like 5%, it is where the rate sits compared to its ‘natural’ setting, so it meant monetary policy was looser than it looked.

Congressman Sean Casten has a thread that explains some issues with banking regulations and the ‘inflation reduction act.’

The way the IRA works is that it declares some forms of investment related to climate ‘good’ so you get tax credits for them. Can you feel the inflation reduction? So that’s great, says Sean, because it means for every dollar in tax credits given out, you generate several dollars in investment activity. We pay $2, industry puts up $10 and we get $10 of windmill if and when it passes environmental reviews and isn’t stopped by the Jones Act.

Sometimes there will actually be a profitable windmill where they put up the same $10 they would have anyway and pocket the $2, but hey, that’s life, and they might do it bigger and faster. Not an obviously crazy strategy.

The problem is that the payment is in the form of tax credits rather than in the form of money. That means that if you are making money, you get paid money in the form of owing less money. But if you are not making money, and presumably need the money all the more, you get nothing. That’s by design. They could have written checks instead and didn’t.

Why didn’t we? Because a certain Senator threw a hissy fit over how it looked:

Sean Casten: Postscript because a few people have said that we fixed that with refundability /​ direct pay. The House version did that – but a certain Senator substantially limited its availability in exchange for his vote. Here’s to tax code (and Senate) inefficiency!

The good news is that banks can get you out of this. The bank invests in the project. As payment, instead of taking money, they take the tax credits, which are money to the bank because the bank owes taxes. So by rerouting banking capital to these projects, we allow the money we gave as tax credits to turn back into money, so everyone involved can feel like they kind of didn’t spend it, and it is only moderately convoluted.

But there is a problem. To do this, the bank must invest capital. We worry when banks invest capital, bank runs and solvency and all that, so we impose capital requirements on the banks before they can reroute our money that isn’t money back into money.

And the Basel III draft rules for how this works say that energy investments are four times ‘riskier’ than housing investments. They do this because there is greater risk in energy projects, much of it due to all the environmental and other regulations that could sink the project. And we are forcing the bank to take on that risk in order to facilitate the tax credit transfer, so it needs to account for that.

Oh no, Sean warns us. If we account for this risk by measuring it accurately, this will cripple the ability of banks to provide the capital, so we won’t be able to reconvert the tax credits. All because of this ‘oversight.’

None of this is an oversight. It is the result of negotiations and deliberate decisions. It would all be deeply funny if the stakes were lower.

Crypto has this issue where people keep getting their crypto stolen.

Crypto also has the problem where crypto people treat this as a marketing issue.

Approve infinity strikes again.

Do you think the user who just lost $4.4million will stay in crypto? Won’t he just sell everything and hate crypto after? It is so irresponsible to build on ERC-20 token standard, but with the current EVM, all token standards will fall to the same problems.

I say the responsibility here is not to the reputation and adaptation of crypto. It is to your users, whose money you want to not be stolen.

Nothing I say is ever investment advice, but we may have spotted Patrick McKenzie giving actual investment advice, and it is the best advice:

Patrick McKenzie: Almost all investment advice is written for people who cannot action the strategy “Choose to earn more.” My investment advice for most geeks begins with “Choose to earn more” and underlining that a lot, because NPV of your career and any optimization of that >>> your $ capital.

Read “cannot easily move the needle drastically” for “cannot choose to earn more” in above. A schoolteacher doesn’t have a static income but they don’t have nearly the dynamism of options available to the people this advice is for.

Thread occasioned by someone who asked for advice given particulars of personal situation which they felt rhymed with my life story.

In the my life story version, best investment in 2010 wasn’t Chipotle even though that was great. Best investment was quitting $40k salaryman job.

I strongly believe this as well, and have acted accordingly. Do something reasonable with your savings, there are various low-fee broad based ETFs available as a baseline option, and then focus on what matters. This holds until you have an extraordinarily large amount of savings relative to potential future earnings.

He also notes that a lot of people who believe that they need to worry about someone draining their bank account, and for the bank to refuse to fix the problem, whereas this is exceedingly rare. It is indeed weird that it is rare, and that we write our account numbers on every check and anyone with the account number can initiate arbitrary transfers out of the account. Somehow we do that, and we have a system on top of it that almost entirely prevents this from going wrong. It still baffles. And yeah, I’m still going to try to avoid putting my account number on various computer servers.

Pat Reginer: When I was in college someone stole my checkbook and used it to clear out my bank account. And then the bank… just gave me my money back. This has informed my intuitions about crypto.

A bold strategy, Cotton, let’s see if it works out:

Patrick McKenzie: The charmingly American healthcare experience of receiving a bill for $89 from a medical office you don’t recognize in a state you don’t live in for a service which sounds plausible but not actually remembered and wondering: scam, data entry error, or actual real bill?

So then you call them and of course that doesn’t work because why would a phone number on an invoice saying “If you have billing questions please call us.” actually result in reaching a human who can answer billing questions.

In Japan that would move the probability far, far towards “scam” but my general feeling is that it moves the probability precisely zero in America.

Anton: I stopped paying any bills that came by mail over a year ago and it’s had zero consequences. Any mail that isn’t obviously personal (hand written, addressed to me, from someone I know) immediately goes in the trash, i don’t even think about it.

“they’ll send it to collections, they’ll hit your credit score” – urban legend, never happened. “important! retain for your records!” – in the shredder with you, then the trash

I have explained to the mail carrier that they’re just creating waste but she refuses to listen to reason.

Anyone can send anyone else a bill for any amount, for any reason or no reason at all. If you don’t pay, they can keep sending the bill and potentially involve collections, again with or without any real reason to bill you for that amount. It is a strange system, or complete lack of a system.

In practical terms, Anton seems largely right. When you see a paper bill, if you do not think it is legitimate, and you ignore it, mostly all that happens is they keep sending you paper copies of the bill. There are exceptions if the size gets bigger, but mostly as far as I can tell they end up writing it off. Often they are ‘making the bill up’ in the sense that you did not agree to pay that amount, and sometimes it is entirely fake, and other times they also billed your insurance and paying the bill would be deeply stupid.

Meanwhile, every legitimate service I use that is not medical, to my knowledge, will bill me only electronically. Makes you think.

Tyler Cowen warns that with fertility on the decline, this could be the last chance for many countries to get rich. If they wait until their populations are in decline, they will face too many headwinds. The obvious response is that AI will change all that, whereas he only mentions AI as making it harder for low-wage economies to offer basic services such as call centers, which seems like such a minor part of the changes coming.

What frustrates me whenever I see such talk is that Tyler emphasizes that the causes of the trend, which he cites as reliable birth control and freedom for women, will not and should not be reversed. But then he does not call for other options or speak of potential interventions, instead he presumes this problem will go unsolved. There is a hell of a missing mood when you warn of countries failing to get rich, when what you are actually warning about is a dramatic and rapid fall in their populations.

At the Movies

Scott Sumner movie reviews for 2023 Q4. Such different worlds we live in. I’ve seen two movies here, Matchstick Men and The Sting. He given Matchstick Men a slightly higher rating, which is bold, but I suspect he is correct. I notice I am much more inspired to watch recent picks, and expect to enjoy similarly rated ones more.

For my own movie reviews, I have decided to try storing them at Letterboxd, with 10 movies so far. I am not claiming to be objective or correct in the way Sumner is. I am going to punish you if the movie is too slow developing, or is not pleasant to watch, although great is still great.

How I’m thinking about the scale:

55 is ‘drop what you are doing, see this and I will answer no questions’ and the only movie of 2023 that clearly qualifies is Across the Spiderverse, I think Barbie is my #2 and on the edge between 4.5 and 5.

My ‘Must See’ threshold is if something gets 4.5/​5 stars, ideally this is also ‘see this and I will answer no questions’ but you don’t need to drop what you’re doing.

I think it is typically a good decision to see anything 3.5/​5 or 45 as well. 35 is either inessential but fun, or has value but also downsides, and could go either way. A 2.5/​5 means this is a subpar product but in the right mood or with a reason, and no better options, sure why not. A 25 means serious issues but there’s something there and it isn’t automatically a mistake. Below that, there isn’t, what are you doing, stop.

I notice that there are kind of two tracks, the ‘this is trying to be entertainment’ track and a ‘this is trying to be art or otherwise do something’ track. It is not the comedy/​drama divide, although that is related. It is also related to Hollywood/​independent, but again not the same and I can think of exceptions.

Of the 10 I saw recently, there were three excellent films that each got 4.5/​5, and I can recommend them to any adult reading this: May/​December, You Hurt My Feelings and Poor Things. I also gave 410 to Godzilla Minus One. I was relatively low on Anatomy of a Fall at 3.5, although I appreciated seeing a very different system in operation, and I was an outlier in the negative direction on Saltburn, which got the only 2.