Ai-Generated

Back in January I argued that AI parallelism turns clean architecture into a forcing function — coupling now has an immediate cost in blocked parallel work. After three more posts living inside this workflow, I want to come back to that argument and say what I think I underestimated. The economic shift is bigger than parallelism. AI has quietly collapsed a whole stack of tradeoffs that software engineering treated as fundamental, and the implications run further than “you can run more agents.” ...

The previous post was about when to go up a level. This one is about the operational layer underneath that decision: which model to use for which kind of work, and the harder discipline that makes the whole thing actually work — being willing to throw away expensive output. The two topics are connected. If you’re not willing to discard, your model choices don’t matter, because you’ll end up patching whatever the cheapest model produced regardless of how wrong its foundation was. ...

The same Sam means one thing at eight and something entirely different at fifteen. The risk curve is not flat across childhood. The social stakes escalate as a child grows, and the escalation is not folk wisdom — it is documented. What happens to the peer world of an American teenage boy between eleven and eighteen changes the calculus completely. Romance becomes the social currency, 11–14 Between roughly eleven and fourteen, romance moves from the periphery to the center of adolescent social life. It comes to dominate the internal world and the conversations with friends; ethnographers and developmental researchers describe dating and romance as one of the organizing principles of adolescent peer structure, not a side activity within it.1 The peer world physically reorganizes to match: kids stop socializing only within their own sex and begin forming mixed-gender groups that bring boys and girls together in joint activity.2 ...

When I was a kid I had a friend who lived nearby. He wasn’t Jewish. He was, by any measure, a good kid — kind, easy to be around, more mature than his age. We played the way children have always played. Hide-and-seek with walkie-talkies and some elaborate set of rules I no longer remember. Pretend games. Pulling bamboo out of the strip behind his backyard, which for some reason felt like the best thing in the world to do on a given afternoon. ...

Relocate the question The primary question was never “should my child have non-Jewish friends.” That question can’t be answered in the abstract, and trying to answer it in the abstract is the error. The real question, the one that determines outcomes: What is the child’s home-based identity actually built on? Because that is what decides whether outside exposure is a minor wind the child leans into and walks through, or an existential threat. Same friend, same Sam, same team — the meaning of the exposure is set almost entirely by what the child is bringing to it from home. ...

The childhood friendship — the neighbor, the walkie-talkies, the bamboo — cost me nothing Jewish. The wrong lesson to draw is “non-Jewish friendships are fine.” The right one is narrower: that particular kind of friendship was fine. “Non-Jewish friend” is not one thing. It names several structurally different situations that get lumped under a single phrase, and the lumping is exactly where the bad thinking happens. A parent who has decided the question in the abstract — for it or against it — has already made the mistake. The situations differ enough that they deserve different answers. ...

In the first post I argued that responsibility in AI-assisted development isn’t line-level review — it’s strategic oversight at the right altitude. This post is about what that means in practice. Specifically: how to recognize when the right move is to stop pushing at the current level and step up a layer. This is the single most useful pattern I’ve found, and it shows up at every scale. The Pattern Whenever something isn’t working — a bug AI keeps failing to fix, a feature it keeps implementing wrongly, a plan that keeps generating awkward code — the wrong response is to push harder at the same level. The right response is almost always to go up one. ...

Before You Drop $500 on a Switch 2, Read This The Nintendo Switch 2 is a great machine. But is it $500-better than a $45 handheld that plays the entire NES (Nintendo Entertainment System), SNES (Super Nintendo), Game Boy, Genesis, and PlayStation 1 libraries — or a $249 one that handles PS2 and GameCube? A Little Story A few weeks ago I went looking for something my kids could play classic games on — something I wouldn’t mourn if it got dropped, lost, or used as a frisbee. ...

A few months ago I wrote a post arguing that AI parallelism would force developers to take clean architecture seriously, because coupling now has an immediate cost in blocked parallel work. I still think that’s true. But I want to start this series with a confession: the elaborate workflow I built on top of that insight is mostly not what I actually do anymore. This isn’t a retraction. The economic argument holds. What I got wrong was the workflow that I thought followed from it. ...

The Sky Arena: Why Air Shows Are a Moral Anachronism We’ve Somehow Normalized We banned gladiatorial combat. We reformed NASCAR until the deaths stopped. Air shows cannot be reformed — they can only be banned. There is a particular kind of moral blindness that only becomes visible in retrospect. Future generations look back and ask: how did they not see it? How did they sit in those stands, eat their funnel cake, and watch without flinching? ...