Nathanael Yellis

In Arthur Brooks's recentpodcast about leisure, he talked about taking it seriously. We don't relax well because we don't relax purposefully, we don't devote ourselves to it. We're either all-on, for professional work and life logistics, or we're all-off, for sitting on the beach or Netflixing an evening away. The podcast initially struck me as classic "type A": the kind of people who want to plan to the nth degree a vacation, in order to see everything. While Brooks certainly tends in that direction, I think he was advising something different: a true understanding of leisure.

After writing about clowns and books and skiing, I'm returning to the reason you read this email: hot takes on AI.

The other day I had this shared experience with a room full of people. Maybe two hundred of us started a day-long corporate meeting. You know the drill: we're there to hear from some leaders about our tactics, learn a bit about how we should be working, and to find the camaraderie missing on endless Zooms and Looms. Maybe it was due to global travel's time zone chaos or just the way these things go, but midway through the first morning, the room was lifeless. There was so little energy, even genuine laugh lines weren't really hitting (one panelist rightly accused the others of "blathering on for so long I forgot the question" to only a few chortles). That was our shared experience: a kind of lifeless lack of energy despite a few hundred gathered people. My challenge was that my panel was next.

There's no better travel accessory than a book: works anywhere, requires no accessories, and rivets the mind. And, while I'll countenance alternate points of view, the size of mass market "pocket" paperbacks is the best form of book. They are four and a quarter by seven inches, fit easily in your coat or blue jeans pocket, and hold your attention for hours. I constructed the shelves of my office bookcase to be just over fourteen inches tall and about ten inches deep to ensure I could quadruple stack pocket paperbacks: two rows high, two rows deep. When I prepare for my flight on Sunday, I'll go to those stacks for a few new (to me) volumes and maybe one old one, for comfort. Everything else will be in my checked bag.

If you ask the wrong kind of person, then you'll hear that this winter has been bad. Real bad, according to them. You see, up here in New England, snow has fallen, multiple times, in such great quantities as to be measured by the foot, and it has stuck. On the North Shore, we haven't seen our gray-brown yards in a month.

It's February, 2026, and us Ice Coasters are having a ridiculous winter. The coastal snow fell and fell again and we had weeks of cold, keeping our powder fresh and plentiful. Among the long-dreamt lines I've skied now is, at the end of Mountain Road in York, Maine, Mt. Agamenticus.

On the train home the other night, I listened to yet another interview with Dario Amodei, the thinking person's AI hype man. At the very beginning of the interview, he gave a little background on himself (neatly omitting stints at Baidu and Google and OpenAI) as a biophysics degree-holder who'd previously done research in microbiology. I thought more about this as he described the monitoring his company does of its LLMs by talking about seeing their "neurons lighting up" in their "brains." It's not the worst analogy, but it illuminates a bit more about the observer (biology guy) than the observed (computer programs).

Today's reading gives you two troubling trends. The first piece below argues, fairly compellingly, that we're all becoming the same. The second wonders if the human species might've numerically already peaked. Both authors, one from data and the other from taste, are not-so-subtly telling us that what's happening isn't good. Both critiques rely on an unfavorable comparison the present has with the past.