So, I've been thinking, for a long time, that a control system might exist in political economy. People like being wealthier, but as they get wealthier, they want to use higher and higher percentages of their income to build comfort and stability. Comfort is a superior good! But they would also expect their governments to do the same. More generous welfare programs, less means testing, higher cutoffs, reduced incentives across the board to make sure things are comfortable for people at the bottom end of the sort. Likely the only thing that would slow down this expansion is if it truly hurt economic growth – so we ought to expect welfare states to grow at least until that happens, and then to find an equilibrium around the lowest acceptable growth rate, which may very well be 0%, depending on the median voter. And if people are wealthy, and that's a true preference, I'm not sure how harshly we should judge. Of course, there is little academic literature I could find about this scenario, but I did find this exploration about how wealth sigmoidally relates to export diversity (and therefore probably how well matched a country is to the people's individual comparative advantage). There are plenty of human-satisfaction improvements to be made, even if economic growth is zero. But I hope that doesn't happen, not until we've converted the local galactic cluster to fuel for our civilizations.

I think (and I can't possibly be alone here) that people have completely lost touch with the elite level of ambition in modern society. This is a helpful guide. I think it's a bit worrying to see comments about how building a web-based IDE is the most ambitious project ever. And prominent leaders in Silicon Valley echoing that they don't think it's absurd. I'll make this claim: I've yet to see an ambition that puts it beyond Lotus 1-2-3. There's something very strange going on in society.

Least Squares is bad. Statistical analysis has always been a bit of an art that used science as opposed to a science, but this is just bad craft any way you slice it. The world is awash is loss functions. Find one that meets the needs of you and your family.

In the "words used to have a small amount of meaning" department, this absolute stunner is just about the most media-illiterate thing I've ever heard. I'm linking to the very funny joke made about this instead of the thing, because we shouldn't give SEO juice to fools and those to pay attention to them. As a side note: if everyone with power and influence agrees something needs to change with the First Amendment, and historically the First Amendment has been critical only for securing the rights of people who wouldn't be able to speak at all, how is the obvious Marxist analysis not happening at all? Am I crazy or is it because "Team Makes-Marxist-Analysis" is one of the powerful players seeking to silence others? Is this a topic we can leave alone and hope Congress doesn't suddenly find enough bipartisanship to be truly destructive?

System size is one of the more important ways to filter results we get. Local businesses being crushed by lockdown produced an exogenous shift to larger systems, particularly those large enough to shift rapidly to online order and delivery-base fulfillment. The next cohort of small businesses will be smarter from the very beginning.

So, some political cartoonists seem to be having a hard time. The Dilbert dude sort of lost his grip years ago, started saying the world is being run by hypnotists, that sort of thing. And now we have this, the strangest Christmas themed political cartoon I've ever seen. Took me minutes to even decode what point he's trying to make. Strange, hilarious, but, I mean, can someone make sure he gets the help he needs?

Please do not spread lies about the terrible bailout bill. It's not great. But it's not as bad as some people are saying. Or, rather, the problems are more systemic. Hopefully we can just get vaccinated quick and stop bothering with all this stuff.

Here's some context for New Years Resolutions. Focus on what matters. Very few things will matter universally, which means you'll be actually introspecting. That's good.

I've been concerned with an acceleration is what I can only describe as critical failures of reading comprehension. Here is an example from someone being reasonable and not having even a self-selected audience understand pretty straight-forward stuff. Here is how bad it gets when someone's trying to explain something even slightly complicated (not all that complicated). Have we lost the patience to engage with others charitably?

I think a lot of COVID risk communication has been exceptionally bad. I've hesitated to say things online, because I think we should all be reasonably cautious about this, but in person, I've... certainly said a lot. It's hard to avoid strong feelings about this. But here's some thoughts about why current public communicators are making deeply wrong mistakes about how to do their jobs. It's a bold claim, but I think it's correct: we've been damaged by lack of competent and trustworthy leaders, even Fauci.

There is a subset of prestige journalism that spends all its time trying to paint tech companies in a bad light. It's very odd. Here's an example that seems pretty strange, for almost any human goal except "hurt big tech because online ads ate our lunch and we can make fun of nerds".

If you're interesting in producivity, you probably spend way too much time on it. But if you'd like to do more and haven't spent time thinking about it, I strongly recommend this interview of Sebastian Marshall. He's pretty thoughtful about stuff like this and his business allows him the incentive to share his insight.

We should probably change how to talk about probabilities to essentially force people to be Bayesian.

This interview was pretty interesting, and not something I knew anything about before. Clarence Thomas' life story is pretty amazing, and while that's a small part of the interview, the rest is amazing too. Thomas gets serious points for professionalism and scholarship.

Good luck next year, everybody.