I have now had to reassure three people that I’m not in fact ghosting them or giving them the silent treatment, so let me put this on broadcast: If you’ve been wondering why you haven’t heard from me directly in a while (apart from these newsletters, I mean), it’s not you.

What’s been happening is that I’ve been less and less able to get the words out, and this past year it got to the point where even responding to a text was almost too difficult to contemplate. Yes, I know what led to it; and yes, I’m working on it. Cards and letters are just starting to trickle out again, but slowly: it may be months before everyone I owe a letter to gets a letter.

Blog posts and newsletters have been less affected than personal messages, but you might have noticed that six-month newsletter gap last year. And to be honest I’ve never been great at reaching out or responding to messages: for decades I’ve had disclaimers on my websites’ contact pages warning that I’m absolutely terrible at replying to emails, and those disclaimers weren’t for nothing.

Truth be told, words have never been easy for me. On a good day I can write a 3,000-word article in an afternoon. On a less good day it can take me three or four months to write a book review half that length. The median is merely a slog. You’d be surprised how long these newsletters take to write.

For someone who styles himself a writer this may seem like an odd admission to make, that writing is hard. But the point is not whether words are easy, it’s whether they are, in the end, good. One can’t judge a concert by the hours of practice put in. Still, the days I struggle over a single paragraph are the days I envy writers who can sustain thousands of words a day.

Nowadays, people who have trouble getting their words out will frequently resort to large language models like ChatGPT. It occurs to me—and I’m aware this may not be an original insight—that generative AI is for people who feel they have to create something, but don’t actually want to do it. Examples of such work are not hard to come up with: job application cover letters, corporate writing. The kind of writing that people pretend to read, so there’s a kind of symmetry involved in pretending to write them. They’re seen as hoops to be jumped through, hurdles to clear—the quality of the product itself is beside the point. It’s a means to an end. You’d think they’d notice the problem when making a living as a writer is the end, and pumping out AI slop, whether to flood Kindle Unlimited or a magazine’s slush pile, is the chosen means, but apparently not.

A related problem is when creatives resort to AI to create ancillary material. They’ll do their own writing, but when it comes to, for example, illustration, they have no qualms resorting to AI-generated imagery to provide blog posts and articles with featured images. What confuses me is the disconnect: artists in one field who’d no doubt be horrified at the use of AI in their own field who have no qualms about using it when another field is involved. See @magicmooshka’s post about this:

recently my friend’s comics professor told her that it’s acceptable to use gen AI for script-writing but not for art, since a machine can’t generate meaningful artistic work. meanwhile, my sister’s screenwriting professor said that they can use gen AI for concept art and visualization, but that it won’t be able to generate a script that’s any good. and at my job, it seems like each department says that AI can be useful in every field except the one that they know best.

You can tell what’s important to the artist and what’s not by what gets done with AI: for example, an orchestra using AI-generated clips during a concert. It’s corner-cutting: you use AI to generate something you’d otherwise have to licence. (Of course, generative AI is built on plagiarism, but no one actually cares about that. In the good old days, they’d just steal someone else’s photo.)

(The one that really baffles me is the frequent use of AI imagery by typewriter hobbyists, because if anyone should be a bunch of Luddites, it’s these guys. But again: AI signals what they don’t care about.)

#

Annals of ultracrepidarianism

Speaking of generative AI: push an AI shill advocate to name one thing it does well, and they’ll usually mumble something about how it’s good at producing code and improving programmer productivity (it doesn’t, but let’s leave that for now). I can’t help but think that this explains, at least in part, the tech industry’s fervour for putting AI in everything: because being good at coding is everything to them. This lot seems to take as a given that writing code is the pinnacle of human achievement (it isn’t, but let’s leave that for now), so if AI is good at code, it must be good at everything else, because everything else is way less impressive and of less value so it ought to be a cinch. QED, or something.

#

‘Justice, truth, and the value of a single human being’

Recently stumbled upon Judgment at Nuremberg (1961) on streaming. I hadn’t seen it in something like 35 years—not since I was still in high school, I think. I remembered the scenery-chewing from the stellar cast (it says something that William Shatner’s performance in this thing was comparatively restrained), but I’d forgotten, or more likely missed at the time (I hadn’t gotten the history degrees yet), the point this film was making.

A fictionalization of the Judges’ Trial, Judgment at Nuremberg is a morality play that explores the question of Germany’s collective culpability for World War II and the Holocaust. But this film did not treat it as an academic question. It was very much a present concern: when Judgment at Nuremberg came out, the Judges’ Trial was less than 15 years in the past; and as the end card revealed, every defendant convicted in the subsequent Nuremberg trials (which included the Judges’ Trial) had already been released. On this viewing it was clear that the film was furious about that. The indecency of compromising the basic principles of justice in the name of geopolitical expedience (i.e., the Cold War), was something this film full-throatedly called out, with Spencer Tracy as its mouthpiece.

I went into this rewatch thinking that this movie might be an artifact of its time. I finished thinking that a movie about the aftermath of a country gone mad, debating the culpability of the functionaries who enabled the regime by going along to various degrees, not just its worst and most enthusiastic perpetrators, and what to do when meting out justice on such people leads to hand-wringing about how impractical or complicated it would be, may well be very pertinent in the very near future. (The nearer, the better.)

#

Miscellany

• Special dyslexia fonts apparently have no meaningful benefit: Adrian Roselli and Youki Terada look at the studies.

• In many reptile species, sex is determined by the temperature at which the eggs incubate; Scientific American asks how global warming is going to mess with that.

• Water monitors are thriving in Bangkok: “No formal census exists, but estimates suggest thousands of water monitors now live across Bangkok. The Thai government has made attempts to reduce their numbers, including a large-scale removal from Lumphini Park in 2016. Yet today, the reptiles are more visible than ever.”

• India’s snakebite crisis—some 50,000 die of envenomation each year—is exacerbated by poor access to medical care in rural areas and a lack of antivenom for dozens of snake species.

• The Walrus has an excerpt from Peter Kuitenbrouwer’s history of maple syrup in Canada that looks at how Quebec maple syrup producers broke a Vermont kingpin’s hold on the industry in the 1920s and 1930s.

• Janus Rose is rediscovering cassette tapes. “There are lots of advantages to the cassette lifestyle. Unlike vinyl records, tapes are compact and super-portable, and unlike streaming, you never have to worry about a giant company suddenly taking them away from you.” James Cook encounters a modern cassette player for the first time (it comes with a pencil, which if you’re old enough you know what that’s for, and also Bluetooth).

• Seiko made a $200 watch that is also a metronome (but only for a couple of years). It’s bonkers and amazing and I started to want one until I remembered that I hate using a metronome. Originally Japan-only; there are lots on eBay.

• Gary Graffman, the American pianist who pivoted to the left-handed repertoire and teaching after an injury, has died at the age of 97.

• Maurice Ravel’s Boléro performed on homemade 8-bit instruments. (Thanks, Jessica!)

• I am the very model of a miniature tyrannosaur: the Nanotyrannus vs. juvenile T. rex debate, Gilbert-and-Sullivan style.

• Owls in Towels, a Bluesky feed of owls wrapped up in towels while in wildlife rehabilitation.