This newsletter finally has a real title

It's about good things in life and how to find more of them

Short updates:

1. Hooray for congestion pricing in New York! Making the tolls more expensive when you expect higher traffic volumes is basically the only way of “solving” traffic¹, and it’s been around for a while in cities like London, Singapore, Stockholm, and Milan.
   
   I’ve personally been for congestion pricing since at least 2017, when I tried to do a data project that looked into the possibility of congestion pricing in Chicago². I’m glad it’s gotten started in an American city because we don’t learn as much from other countries as we should, and hopefully it spreads to other cities that need it.

2. Ballaké Sissoko (kora) will be playing Old Town School of Folk Music with Vincent Ségal (cello), Emile Parisien (saxophone), and Vincent Peirani (accordion) on March 26. This one sneaked up on me because it’s not listed as one of Ballaké’s tour dates on Spotify or Songkick. See Peirani’s tour dates here. In North America, they’re also playing in Pine Plains (NY), New York City, Somerville (MA), Minneapolis, and Ottawa.
   
   Curiously, Parisien’s website shows additional tour dates in Seattle, Vancouver, and Victoria. Not sure what’s going on there.

This Substack has been called Satoru’s Newsletter since 2022 when I started writing it. While that’s a good way to keep anything and everything up for grabs, it doesn’t give anyone a reason to be interested unless they’re interested in what I have to say to begin with. So I’d been meaning to come up with an actual title, something that describes, or at least suggests, what I’m writing about.

The new title is Satoru’s Greedy Algorithm.

I’ll explain what I’m going for in this post, and I’ll keep it for the foreseeable future unless someone can convince me that it’s terrible.

What has this newsletter been about?

I do want to make it clear that the new title doesn’t mean I’ll write about vastly different topics than before. It’s more a way to tie together the topics that I’ve already been writing about into some sort of a framework.

So what have I been writing about?

1. One major theme is art appreciation, broadly.

   1. Music has been part of every post. Even when it’s not the main topic of the post, I have a What I’m listening to now segment at the end³.

   2. Food has often been the main topic of my posts, the most popular being this one on ramen in and around Chicago (with two followups).

   3. I’ve also talked a few times about art exhibits that I saw.

2. The other main theme is trying to figure out how things work, and how they may work in the future.

   1. Many of the posts about Chicago and other cities try to figure out why some things in these cities are the way they are.

   2. I have a couple of posts explicitly about AI⁴.

How does the new title capture these?

Obviously, I’m writing about these things because I’m interested in them. But why am I interested in them? At the risk of over-intellectualizing something that could just be my personal biases, here’s my attempt to reorganize these topics into the words “greedy” and “algorithm”⁵ (you can safely skip to the next section if you only care what will happen with this Substack):

1. Greedy: A very short answer is that “greedy” is obviously a good word for a Substack that talks about food a lot.
   
   I also have a very long answer.
   
   Increasingly, I’ve been drawn to 2 ethical traditions—utilitarianism and Epicureanism—that teach you that it’s morally a good thing for people to seek pleasure and avoid pain, to try to be happy. Most people would agree that they like having pleasure rather than pain. But is it the right thing to do, to prioritize it? You could probably think of situations where enduring pain seems like the right thing to do. So this isn’t an obvious thing to agree with, but I am agreeing with it.
   
   Hopefully more to come on these philosophies, but my writing on food and music is about having good, pleasurable experiences. Having good experiences is a surer way of being happy than having more stuff. And even better, food and music are two arts that can be very communal—the point is often to enjoy them with other people. That builds relationships and community, which makes them even surer ways of being happy⁶.

   
   One caveat about focusing on good experiences is that when you start comparing yourself with others to see who is having the most fabulous experiences, it tend to go badly because then you’re in an arms race. So I want to remove the status game from these experiences as much as I can.
   

   The academic term for the idea that morality should be based on pain and pleasure is ethical hedonism, which sounds bad but it’s just what it is. There’s inherently an idea here that some actions that may be labeled “sinful” by others may actually be ethical, suggesting “greedy” as a self-deprecating label⁷.

2. Algorithm: Another thread that I’ve been exploring, especially with respect to AI, is the tension between the following beliefs that I hold strongly in my head:

   a. Humans are awesome: Cities are awesome. Food is awesome. Music is awesome. A large fraction of what I value in the universe was made by humans.

   b. Humans do not inherently have a special status in the universe: In particular, there’s nothing supernatural about the human brain.
   
   For the first belief, I hope it’s clear that I wouldn’t spend my time writing about these things if I didn’t think they were great.
   

   As for the second belief, when we feel that we’re making a decision or a judgment freely, our whole body is simply following the laws of physics. It’s really hard to sneak free will back into this picture without sounding like a teenager who wants an exception to the rule. My suspicion is that human thought is deterministic to a large degree, with some randomness thrown in, and that free will is an illusion that has been useful for us⁸.

   
   So I’m claiming it’s fair to compare human thought to algorithms, which are well-defined procedures that turn inputs into outputs⁹. This thought connects to so many questions I have about what’s good in life: How do we decide what’s good food/music/etc.? How do we find these good things? Why can/can’t computers make good things?

   
   When you see the word “algorithm” in non-technical settings these days, it’s usually when someone complains about things like social media feeds or AI, where they feel that there’s an opaque and unaccountable piece of computer code that is controlling some aspect of their lives.
   
   I guess the title is doubly self-deprecating.

Now, the people who have studied algorithms are screaming at me: ‘You started with the phrase “greedy algorithm” and worked backwards!’

And they’d be correct. Greedy algorithms are algorithms that always go for the best immediate results—it’s not so much greedy as shortsighted¹⁰. I’ve always liked the phrase as colloquial-sounding terminology, and it turns out that I can make a long-winded argument for how both “greedy” and “algorithm” apply to what I’m writing.

So what does it mean for what I’ll write?

Here’s what I expect:

• I’ll still write about 1 post per month

• Just about everything that I’ve written so far can fit into the themes I presented above. What might change is the focus, and one thing I’ll try to write more about is my process for finding whatever it is that I enjoyed.

• One topic that I’ve been meaning to write about, but haven’t yet, is parenting. How having a child (or two) can be a source of joy (or not) seems like an obvious topic.

What I’m listening to now

In a funny coincidence, I’ll be talking about the same artist as in my first post for Satoru’s Newsletter. Then, I talked about De Todas las Flores by the Mexican singer Natalia Lafourcade. She played at Carnegie Hall in 2023, and the live recording came out near the end of last year.

The first half covers most of the songs from De Todas las Flores. In the studio album, I especially liked 2 of the lighter songs, “El lugar correcto” and “Caminar bonito”, but the live band really nails the more sombre songs, like “Pajarito colibrí”, “María la Curandera”, and, of course, “Muerte”.

The second half is older material. She had some truly special guests for this show: David Byrne reads an English translation of the lyrics for “Muerte” (basically the end of the first half), then Jorge Drexler—the Uruguayan artist that I talked about in my second post—does an excellent duet on “Ya No Vivo por Vivir”, and Omara Portuondo of Buena Vista Social Club fame is in a couple of songs.

She’s probably the most popular artist that I’ve covered in this space, if we look beyond gringo culture. I consider her one of the greatest singers of my generation.

Final note: she just announced a tour. She’ll play most major cities in the US, though somehow not a single show in Southern California. I will try to make her Chicago show on 6/14, but that may conflict with a Japan trip that we’ve been hoping to take.

1. Increasing road capacity does not work in the long term, because of something called induced demand—people start driving more, and more people start driving there, when there’s more capacity.

The way I like to explain this whole idea is that traffic is a tragedy of the commons problem. Because the road is usually free, people use too much of it. And when too many people are on the road, everyone loses out because they can’t move quickly. When you add one more car when the roads are busy, you are slowing everyone else down—“You aren’t stuck in traffic, you are traffic.”

Congestion pricing discourages you from imposing that cost on others.

2. My problem was that there wasn’t a great public dataset for it. I mainly looked at the taxi trip dataset released by the city.

3. I usually don’t have the segment when the post itself is only about music.

By the way, I have a Spotify playlist of all the songs/albums I talked about in the segment.

4. As a data scientist with some experience working with deep learning models, I can claim some expertise on existing AI systems. I hate talking like this so this is a footnote.

5. A lot of this thinking happened during my parental leave. It’s an interesting period where you have both so little time (to perform all the tasks) and so much time (to think about stuff).

6. Non-exhaustive list of other communal arts and activities: theater, dance, games (including sports), religion.

7. To be clear, neither Epicureanism nor Utilitarianism (nor I) sees excessive desire for material gain as a good thing. I’m mostly pointing out that there have been criticisms that these philosophies promote sinfulness.

8. The key thing is that our ancestors must have done better evolutionarily if they thought they had free will.

9. We now think of algorithms as being computer algorithms, but this isn’t necessarily the case. Long division is an algorithm, usually done by hand, for computing division involving large numbers without lots of trial and error.

10. One irony here is that Epicureans and utilitarians do see the problem with being shortsighted in seeking pleasure. They would argue that enduring pain for a bigger payoff later is obviously the right thing to do.