https://lukescalone.com/bookmarks/ Links worth keeping, newest first. en-us Wed, 24 Jun 2026 00:19:44 -0500 https://www.publicbooks.org/dear-syria/ lukescalone-bookmark-23 Essay Wed, 24 Jun 2026 00:19:44 -0500 > As weighty as my visits with people are visits to places. So many towns and villages carry legendary histories of rebellion and war, and they are the backdrops of testimonies that I have recorded over the years. Comparing their images in my mind’s eye to the reality before my actual eyes frequently leaves me awestruck. How could Daraya, a town whose revolutionary resourcefulness and resilience are of mammoth proportions, be this small suburb? How could cars get tangled in Homs’s clock-tower square, as if it were a regular traffic circle rather than the site of a turning-point vigil and massacre? How could the dreaded Palestine security branch, where some of my interlocutors had undergone unspeakable torture, be that nondescript building on the highway? How could I walk from what had been government-controlled western Aleppo to what had been rebel-controlled eastern Aleppo, both of which existed in my mind as distinct universes separated by the infamous “death crossing”—which is now just a road? > Syria became a long-distance pen pal to whom I was passionately devoted but never expected to meet. This essay gave me chills---reading the way she navigates the *affect* of space feels something like looking in the mirror. "Oh, you mean the grand ports of antiquity are a leafy suburb?" "What do you mean this great battle is now an unmarked field in Siliana governorate?" I love seeing people navigate the tension between the mundane and the mythic, it's a reminder that the mythic only takes place---can *only* take place---within the basic structures of everyday life. https://timothyburke.substack.com/p/teaching-is-to-writing-like-ministering lukescalone-bookmark-22 Post Tue, 23 Jun 2026 23:34:45 -0500 I share Timothy's views on teaching writing in the age of generative AI. Responding to technical problems with a technical solution simply will not work. Beyond the policing argument---which I find correct---it also, simply, will not work, as tech advances faster than surveillance software ever can. My own solution, like Timothy's, is to turn towards writing in class, by hand. Even homework outside of class should be done by hand, if for no other reason than (in my experience) hand gestures allow us to *think* in ways that typing on the keyboard does not. Slow rigor should come over scale any day. https://www.mcsweeneys.net/articles/ai-economics-for-dummies lukescalone-bookmark-21 Other Mon, 22 Jun 2026 01:17:22 -0500 "Trust me, bro. It makes perfect sense if you think about it hard enough." The deepest issue with transformations brought about by AI are its political economy and it---honest to God---makes no sense at all. Generative AI is really valuable, but the numbers don't work and they can't work. Making matters worse, each new model launch is *more expensive* than those that came before. https://www.publicbooks.org/who-benefits-from-distorting-american-studies/ lukescalone-bookmark-20 Essay Wed, 17 Jun 2026 15:06:25 -0500 After finding this article, I took a look at the original study by Kahlenberg and Lin, and the whole study is suspect. Methodologically, it approaches the subject by classifying academic articles as "anti-American," "neutral," or "pro-American." It then finds that the overwhelming majority of the articles are anti-American with a smaller number that are neutral. The issue, I think, is that the very idea of having a balanced valence--pro-American and anti-American--is neither productive nor helpful. The whole point of academic journals like *American Quarterly* is to be *critical*. When I say "critical," I don't mean "negative." What I mean is that they do not take things as they are, but dig into deeper structures to figure out what might be missed. So, Kahlenberg and Lin give, as an example of a potentially "positive" story, the United States' triumph over Nazi Germany. On the surface, this *is* positive. But, dig a little deeper, and what is there? The Second World War ended with (1) nuclear bombs dropped on Hiroshima and Nagasaki, (2) attempts at de-Nazification that were aborted when the US began to get skittish about the Soviet Union, (3) freezing tensions in the world order, and (4) the import of Nazi scientists to the US. At the same time, the very thing that made the Nazis bad domestically--the construction of an elaborate racial state--*continued* in the United States: civil rights laws were not passed for another two decades, and they themselves were not able to overcome the structural changes the ensured racial discrimination continued across the United States (zoning, redlining, infrastructural development, white flight, and so on). Critical scholars, like those writing for *American Quarterly*, have examination of these processes as their *raison d'être*. The journal doesn't pretend to be a triumphalist organ. The better question, which Kahlenberg and Lin's study doesn't exam, is: "Are they effective, and what can we learn from it?" https://lr0.org/blog/p/arabic/ lukescalone-bookmark-19 Post Sun, 14 Jun 2026 18:30:28 -0500 Phenomenal article on the trials and tribulations of trying to get Arabic type to function on computers. The post itself has some phenomenal interactive elements, where you can see precisely how ligatures form and the way that printing without proper encoding breaks the text entirely. https://www.wsj.com/tech/ai/anthropic-claude-ai-vending-machine-agent-b7e84e34 lukescalone-bookmark-18 Essay Thu, 11 Jun 2026 22:58:34 -0500 The sequel to my previous bookmark, and---once more---I find it so good. It says that the two models being used here are Sonnet 3.5 in Part 1, and Sonnet 4.5 in Part 2. I wonder how more recent Opus or Fable models would compare. And, I really want to put two AI agents in a Slack chat and have them converse. Super interesting. https://www.anthropic.com/research/project-vend-2 lukescalone-bookmark-17 Other Thu, 11 Jun 2026 21:33:04 -0500 The relationship and antics between Claudius, CEO Seymour Cash, Clothius, and CEO Big Mihir are significantly funnier than I ever expected them to be. The reason they're funny, I suspect, is because *I can see a human doing exactly these things*. A badly trained human who doesn't know anything about the world, but a human all the same. The one thing here that might be considered "alien" is the "$527+infinite pipeline" completed over "4 continents transcendent." I have no idea what this means, but I am *curious*. https://www.newyorker.com/magazine/2023/03/06/the-end-of-the-english-major lukescalone-bookmark-16 Essay Tue, 09 Jun 2026 17:16:37 -0500 There's a lot unresolved here, and it necessarily must be so. Heller draws significant attention to the decline in humanities *majors*, but---after reading the rest of the essay---it hardly seems that interest in the humanities is lessening. Students remain deeply curious about the humanities, and they tend to be some of students' favorite courses. However, undergraduates worry more about the job market and their economic prospects than the pleasure of deep, creative engagement with texts and ideas. I suspect that the bulk of this problem is something that cannot be resolved by universities themselves. Administrators are busy chasing after funding, which has declined an enormous amount since the end of the Cold War. So, universities have shifted their attention to working in favor of career readiness and improving student outcomes after graduating. In the process, they have folded humanities courses under other disciplines: history of science tied to B.S. programs, ethics appearing in bioethics and AI ethics, reading and writing merged with communications and media studies. The commonality shared by all of these is the subordination of critical thinking to other practical ends. The way out of this impasse is to reframe the university, as has been done numerous times in its millennium of existence. By the late nineteenth and early twentieth centuries, higher education had become a laboratory for the skills necessary for sincere democratic politics. By the late twentieth and early twenty-first centuries, a source of capacity-building for cognitive labor markets. What, then, is the university for? My stance is that the idea of universities as training centers is too profane. Perhaps I've sacralized this institution, but it is the one institution that I have found where learning is taken seriously for its own sake: some students chase after grades and networking, but it is the sole place I have ever inhabited conducive to sharpening the intellect without any further purpose, and it would be a tragedy to lose it. https://patience.toys/ lukescalone-bookmark-15 Other Mon, 08 Jun 2026 22:15:17 -0500 I found this site through [A Website is a Room](https://a-website-is-a-room.net/) and am so impressed with it. It has little, interactive tasks that train patience. Soft sounds of waves accompany the game, and it's---at core---meditative. https://neustadt.fr/essays/the-small-web/ lukescalone-bookmark-14 Essay Mon, 08 Jun 2026 16:15:49 -0500 Love this piece. I had originally felt discontented with loss of what the website once was (and now is), and I only felt it acutely after launching a "blog." This is a great exploration of the early web and the way it differs from today. I don't love the repackaging of 32-bit graphics and styles, but the philosophy behind it is worth recovering. https://www.wired.com/story/the-painful-truth-about-long-covid/ lukescalone-bookmark-13 Article Sun, 07 Jun 2026 23:14:40 -0500 The most important intervention is near the end of the article: just because an ailment is "psychological" does not mean that it lacks biological sources. Moreover, it seems to be that critics of the psychological perspective are far too defensive---these doctors are not saying "it's all in your head," they're saying something more sophisticated about the relationship between cognition and experience. I believe that, at one point, I suffered from long Covid, but I may not have; it's hard to tell. I remember any climb up a flight of stairs being overwhelming, I remember walking down the street and feeling utterly fatigued, and I remember not being able to sleep at night, feeling jittery the whole way through. The way that one expert here refers to the "fight or flight" reflex being on overdrive rings true to my experience, although mine is hardly the only one. However, the article has a missed discussion. In one definition of "long Covid" (the standing one?), there is a point made that an individual need not have been positive with Covid-19. This is *really* interesting. If it's true, say, that long Covid is psychological and comes out of fight-or-flight reflexes on steroids, it need not follow that it be caused by Covid itself. If I remember right, the age of Covid was a particularly stressful time *everywhere*. In the United States, it coincided with drastically escalating racial tensions, democratic backsliding, a populist quasi-coup attempt, and social estrangement. While we were not always "present" for these events, being locked indoors ensured that mediatized realities were broadcast to us all the time, and there was hardly a good way to "touch grass" and ground ourselves. These spectacles became our lived experiences and we absorbed them as catastrophe. It seems that more scientific work needs to be done---and I really don't think cognitive-behavioral therapy and exercise are as harmful as critics makes them out to be---but it is worth keeping the *full* context of that moment in mind. https://www.thebignewsletter.com/p/a-billionaire-explains-why-american lukescalone-bookmark-12 Post Sat, 06 Jun 2026 22:25:36 -0500 Fascinating discussion of the transformation of the American regulatory industry through the lens of John Malone. The most fertile ground for discussion is in John Malone's question to Mark Zuckerberg about "why he sought to acquire rivals instead of buying smaller percentages of them." Zuckerberg's response is telling: "Well I don't know why I would do that when I can sort of own everything." Malone is surprised by this---by how lax the federal government had become in enforcing anti-monopoly policy. Yet, what is missed is that Malone is part of the reason *why* anti-trust regulation became as weak as it is today. In other words, Malone's desire for how the government might be was quickly outpaced by the transformations that took place around him. There's a missed opportunity here to talk about that further, but, on the whole, this is a well-written Substack entry. https://www.liberalism.org/p/enshittification-despotification-and-the-open-internet lukescalone-bookmark-11 Article Sat, 06 Jun 2026 22:00:38 -0500 Masnick gets the historical trajectory correctly (enshittification and "despotification" go hand-in-hand) and despotification is really about chokepoints---we might call the current age, generally, the rebirth of the chokepoint; see, the Strait of Hormuz. However, I'm not convinced by his conclusions. I agree without reservation that decentralization is the way forward, but federal systems like ActivityPub and Bluesky don't seem like the solution as such. The reason that people flocked toward platforms like Facebook, MySpace, and Google is that they lowered the necessary *capacity* required to use online tools. Each innovation lowered the degree of human know-how to move forward: all of them built on the metaphors implemented through Windows and Mac, and the incorporation of AI tools within (including in search, as Google is doing with Gemini) is not qualitatively different from what came before. A decentralized internet can only work if people know how to use it beyond point, click, and type. Pages and applications have to be *created* and *curated*. If the problem is the algorithm, decentralization---on its own---cannot solve that. Only bringing back the humanity of the web can. The early web had a very high barrier for entry: not everyone knows how to use the command line, write Python scripts, or even produce the most rudimentary HTML and CSS. It seems that the only true solution is to invest heavily in building these technical skills, without cutting time given to humanistic capacities. Perhaps geometry and trigonometry courses, which have *content* used by few but skills used by just about everyone, can be replaced by these deeper, technological capacity-building courses. https://github.com/kdeldycke/awesome-falsehood lukescalone-bookmark-10 Other Sat, 06 Jun 2026 21:26:18 -0500 Fascinating list of falsehoods held by programmers; essentially, things they take to be true, encode into computer systems, but that lead to serious detrimental results. https://www.nplusonemag.com/online-only/online-only/in-the-reality-lab/ lukescalone-bookmark-8 Essay Sat, 06 Jun 2026 01:23:29 -0500 Meta has stopped trying to solve problems and started building things nobody asked for, on the theory that one man's conviction about "the future" can stand in for anyone actually wanting the thing. Nobody is asking for these wearables, but Zuckerberg seems dead set on selling them anyway