In praise of artificial stupidity

When I was looking for a career change after my physics doctorate, one obvious direction was data science. There were bootcamps for people in my position, it was highly paid and highly valued, there were still not as many specialized fully-fledged training programs to compete with, so I looked into it. And to be honest, it just wasn't clicking with me. I wanted to like it, as I did have a scientific background, as well as in computer science and software development, but it was just not my thing. I realized after a while that I wasn't satisfied in finding a model that seems to provide patterns that showed that it knew something - I wanted to know and understand things myself. So I instead went into game development, and particularly, computer graphics.

Years later, it seems like computers understanding things for us has just become more and more popular. Professional computer graphics conferences are full of machine learning algorithms and neural networks for many purposes, with hardware technologies to go with them - one talk at SIGGRAPH 2021 distinguished itself as providing a non-neural graphics algorithm. Moreover, computers are now capable of not only processing things for us, but of generating a lot of creative work, leading to a lot of concern and downright disruption of the lives and livelihoods of the people who do this for a living. More and more it seems like all these techniques and technologies under the rubric of artificial intelligence are on the upswing. Maybe I made a bad bet.

But I, personally, prefer artificial stupidity. Take Newton's second law: F = ma. It says that in the contexts that it explains, nature can be modeled as something fundamentally very stupid: it just takes forces and converts them to accelerations. Even if we model something complicated, like a large bridge, or a system made of up thousands of particles, and are not able to perform all the calculations ourselves, when we tell a computer to perform those calculations for us, we're just telling it to do something stupid faster than us. We still understand what it's doing.

Of course, there are aspects of what a computer does that are not as obvious as they used to be. We are using an abstraction that has become less and less representative of the actual behavior of the hardware. When we tell a computer to calculate things, it might perform these calculations in an order different from that which we gave it, under the constraint that the results are as if it had done them in the right order, but in practice, much more quickly than it could have had it just abided by our instructions. But even these ways of working are understandable. Maybe this will change in the future.

Generally, we are now faced with a lot of "algorithms" that people do not understand, but just accept as the way decisions are made. That's the opposite of what algorithms were always for me, when I was studying computer science - even if clever, they ultimately did understandable things, and if they were doing things in ways that did not make sense, that was a reason to investigate, not to shrug and accept what was given.

Computers are tools for our use. But to be tools, they have to be understood. It needs to be clear when they aren't working - and when they get to set both the results and the standards, we can't do that.

Is episodic television roguelike?

I recently finished the first season of the old Perry Mason show from the 50's and 60s. I haven't watched it before - instead, I came to it after having watched the new, short-lived HBO show with the same title and almost the same characters. That gave me the reverse perspective from people who have watched these chronologically.

The HBO show benefits and suffers from a pattern of prestige remakes of old shows. In the old show you see defense lawyer Mason, his secretary Della Street, and his trusty private investigator Paul Drake, save a client from a murder charge by extreme means as of the first episode. In the new show, the first episode introduces us to Mason as a private investigator, Paul Drake as a police officer being, and Della Street as legal secretary to a lawyer who is Mason's mentor. The season ends with Mason transforming into a defense attorney, Street becoming his secretary, and Drake somewhat more associated with him. Mason's court nemesis from the old show, Hamilton Berger, is much more ambiguous here, and is a mere assistant DA. Lt. Arthur Tragg, the homicide detective and the one usually more closely involved in making a case against Mason's client, is not significant here. Since the show was canceled before it could be completely finished, I don't know how far they would go in reproducing the dynamics of the original show. Since it's supposed to be more closely based on the books than the original, it may never have been planned to reach that status quo at all.

The original show in light of all this is fascinating. It's a truly episodic show. There is no reference to earlier episodes, no characters develop in writing. Each episode stands on its own. They also all have a very similar pattern: we are presented with a non-deadly conflict, having to do with finances, emotion, or both - maybe someone is being blackmailed, maybe someone's heart is broken, maybe someone is being cheated out of an inheritance - and someone involved contacts Mason, and becomes his client. Then someone that was in conflict with the client dies, and Mason is diverted from corporate and real estate law to criminal defense. Mason himself often finds the body, and his law-stretching attempts to balance his duties as an officer of the court and his duties to his client bring upon him the ire of Lt. Tragg, who is not far behind. As more evidence is uncovered, Mason's client looks more and more guilty, until he comes up with an insight that lets him not only avert his client's conviction, but help Tragg and the district attorney, Berger, find the real culprit. Usually this involves the guilty party blurting out a confession in open court.

I watched a full season of this. In fact, many people, from the early history of television, and to this very day, watch shows like this, for many, many seasons. Shows in which the episodes are connected by the characters and the circumstances, but don't explicitly build one episode on the other.

It used to be suggested by the medium: in the time before VCRs, and with reruns and syndication depending on popularity to begin with, you simply couldn't depend on your viewers to have watched several shows in a row in order to follow any kind of serial narrative. You had to make every episode stand on its own, but you also had to make it reproduce the brand you were marketing. So repeating the same episode premise with some variation was a pretty reasonable design decision. Once you knew you had an audience, you could start playing with the medium more. You could end episodes or seasons with cliffhangers, and use those to encourage people to watch next time. You could change the status quo. That being said, you didn't have to.

So I watched 40 episodes (39 from the first season and one from the second). What did I get out of this? The writers did not present me with character growth. If you mixed the episodes up I wouldn't have been able to see the difference. But I changed. My frame of mind about the show changed. As I watched episode after episode, I got to see more and more of these static characters. Instead of seeing how Mason grew into his position, how he formed the relationships he has with his allies and antagonists, I saw Mason in this position, I saw these relationships, under many variations of the same challenges. I saw Mason and Tragg maintaining an almost erotic tension, so delighted to one-up each other, but with a certain moral commonality that meant that when Mason really managed to find out who the real culprit was, Tragg would cooperate with him, waiting in the wings to make an arrest once he'd pressured the guilty party, if he couldn't do it in court. I saw the legal procedures, and possibly understood them better - when it does make sense to object, when it doesn't. I could start seeing patterns of when it's clear that Mason or Burger were going out of bounds intentionally, as a rhetorical point, and when they've clearly been stymied by the other party, or by the behavior of a witness. If an episode did go in a very unusual direction - so far only one got all the way to a jury reading a verdict for one of Mason's clients - it could be used to examine the others in sharper relief.

I think that the process of a viewer of such an episodic TV show is like that of a player of a roguelike video game. A roguelike usually has the same general premise from playthrough to playthrough. However, the details are procedurally generated. In the more strict versions, you don't get to keep anything, so you can't build up a mechanical narrative of change as you play. What you do is learn, from session to session, about the laws of the world, of the patterns of action of the enemies, of the power-ups, of the dynamics of difficulty as you move farther and farther in, or travel farther and farther out. You understand existing things - those which don't change from playthrough to playthrough - better, due to how they interact, or how they relate to each other, through what does change each time.

Arcade games are similar. You play Galaga over and over. The game starts over each time, regardless of how well or how poorly you did in your last run. However, these games usually have a certain progression that is repeatable, so that you end up learning that, more perhaps than anything more inherent about your avatar in the game, or the enemies, or your power-ups. Often you can memorize some of the specific patterns enemies take in specific levels - in fact, depending on difficulty, it might be a requirement to advance.

So I think that (strict) roguelikes might be the best gaming exemplars of what I'm talking about. Might be time for me to try looking at this genre anew.

Trying something different for 2024

Greetings, faithful readers! It's been a while!

The purpose of this blog was to be an informal creative outlet. Sometimes it's worked out. Mostly I've been overthinking posts to the point of not publishing them at all. I have way too many drafts and half-baked ideas, and they're not getting to you. So I'm going to try something different this year.

I will try to set myself a schedule of posting once a week, starting today, but usually on Sunday nights. Instead of requiring myself to have a great piece of writing ready, I'll just treat this as a log, a journal, a commonplace book - somewhere to write down my thoughts with a bit of editing, then share them with the world.

In the past, I have been more prolific when I forced myself to spend October on some daily project. That is a bit much, especially with a day job, but once a week, just getting those thoughts out there, whether or not I have something fancy to share, should be doable.

So, what is up with me? For a while I was reading too many books at once, which was good in that I was making some progress, slowly, but bad in that I was having a hard time getting a clear sense of any individual book, to the point where I feel like I'd have to reread some of them to get anything meaningful out. Once I started committing myself to reading fewer and finishing them, I've felt a lot more accomplished.

I haven't read anything really folklore-related recently, hopefully I'll get back to that, but I've been trying to get at my other interests. For example, I recently finished reading Dungeons and Desktops - The History of Computer Role-Playing Games, 2nd Ed, by Matt Barton and Shane Stacks. Some of the games I've played, but a lot of them I didn't - either way, it was great to get a deep and broad sense of what this part of gaming has going for it. This is a field that was dying when the first edition came out, but has been getting more popular in time for the second, so it was interesting to get a slightly removed analysis of this revival.

I finished reading the second book in my Ervingiad, Encounters: Two Studies in the Sociology of Interaction. Unfortunately, it's another victim in my being spread too thin problem, so I'm now rereading it and summarizing it to myself. So far I've been finding it most illuminating - in particular, the book hinges on the distinction between analyzing certain activities that involve a set of people, and the structure of small groups, with obvious applications to, say, role-playing games.

Speaking of role-playing games, I haven't run a game myself in more than two years, although I have been playing Call of Cthulhu pretty consistently, with occasional dips into other systems. I hope to run games again this year, starting with attempting to do Apocalypse World one-shots. We'll see how that goes. Hopefully better than my one attempt at doing so for Sorcerer.

I've also been working on my French again. And in that context, I've gotten back to attempting to read René Daumal's Le Mont Analogue, which I got years ago. I now sit down with it, go to the dictionaries when needed, resort to Google Translate when all else fails, and climb through. I'm almost two chapters in. It's fascinating.

Aside from these I have a bunch of few-sentence book, film, and TV reviews from the past few years that currently exist as messages on Discord, and which I should collate some time. Incidentally, I've watched a few good movies this year, including Promising Young Woman and In the Mouth of Madness, and read pretty much all the literature recommended for Sorcerer. I might try to collect the reviews and my thoughts and share them in future posts.

Here's to more of me in 2024!

Layered mysteries

WARNING: The following post contains significant spoilers for Raymond Chandler’s novel The Big Sleep, for the 1973 film The Long Goodbye based on his novel on the same name (so probably for the novel itself), and the video game Castlevania: Symphony of the Night.

Weird Circle 31 - Wuthering Heights

"The Weird Circle. In this cave by the restless sea we are met to call out of the past stories strange and weird. Bell-keeper - toll the bell, so that all may know that we are gathered again in The Weird Circle." "Out of the past phantoms of a world gone by speak again the immortal tale: Wuthering Heights."

This episode, first broadcast on March 26, 1944, is an adaptation of Emily Brontë's famous novel of the same name, first published in 1847. The adaptation took great liberties with the original, condensing the significant elements of the plot involving Heathcliff and Catherine, and framing it as a gothic tragedy with supernatural elements. Heathcliff's treatment is undersold, so he seems far more of an outright villain. On the whole, though, it was a very compelling adaptation, and makes me want to finally go ahead and read this classic, and see if I can find the horror elements that have been extracted here.

And so ends my spooky project for October 2022. It has been a very different experience than last year's. I'm not sure if The Weird Circle was the best choice of series for this project, but it did introduce me to a variety of authors and their works, and it was interesting to see a breadth of quality, and to try and think of how this speaks to radio as a medium. I might have a bigger summary post later in the week.

Weird Circle 30 - The Strange Judgment

"Out of the past phantoms of a world gone by speak again the immortal tale: The Strange Judgment."

This episode, first broadcast on March 19, 1944, might be an original - although as I was looking for its origin, I found that a previous episode, The 4:15 Express, was actually based on a short story with the same name by Amelia B. Edwards, which can be found in this collection with other authors on Project Gutenberg, or in this scan of Monsieur Maurice and Other Tales, which is where the story was first published in 1873 on the Internet Archive. I should have consulted resources like this more often.

Back to today's story: our protagonist is isolated at the top of a hill overlooking his former home village, barred from speaking to anyone. The tale is of why: of his success at fishing sardines, at him and his wife welcoming a child, and that child's promise and frustration, culminating in a horrifying conclusion which led him to this penance. Moving, although sometimes inexplicable.

Weird Circle 29 - The Spectre of Tappington

"The Weird Circle. In this cave by the restless sea we are met to call from out of the past stories strange and weird. Bell-keeper - toll bell, so that all may know we are gathered again in The Weird Circle." "Out of the past phantoms of a world gone by speak again the immortal tale: The Spectre of Tappington."

This episode, first broadcast on March 12, 1944, is an adaptation of the short story with the same name in the collection The Ingoldsby Legends, penned by Richard Harris Barnham under the pen-name "Thomas Ingoldsby", and first published in 1840. The adaptation spreads aspects of the third-person narration of the original more evenly among the active characters while reducing their overall number somewhat, making it much more lively for radio, but the legend of the Spectre, the romance, and the mystery of the stolen pants, as well as its conclusion, remain.