Here’s the start of Chapter 7, ‘Naive Intervention’, from Antifragile:
Consider this need to “do something” through an illustrative example. In the 1930s, 389 children were presented to New York City doctors; 174 of them were recommended tonsillectomies. The remaining 215 children were again presented to doctors, and 99 were said to need the surgery. When the remaining 116 children were shown to yet a third set of doctors, 52 were recommended the surgery.
[…]
Let us call this urge to help “naive interventionism.”
Goodreads tells me that I read the book in 2019. I honestly can’t remember too much about it, but that paragraph has really stuck with me. From time to time, when a friend or relative would say something like, “My doctor said I should …”, I’d pull it out of the closet in the back of my head and try to sound smart and say, “Well, you know, there was a study once…” Then I’d fumble the numbers, of course, and I’m pretty sure that multiple times I made it about wisdom teeth and not tonsils, but the point I would try to make is that people whose job it is to do X are biased toward thinking that doing X is more important than not doing X.
And now I’m wondering: is this what’s going on? Is this what’s happening when engineers look at the output of a Sol or a Fable or an Astra and say “it writes bad code, it leaves all these dumb comments”? Bad code? Dumb comments? Really?
Or did it just knock out a feature, end to end, in the 20 minutes you weren’t looking, including frontend and backend changes, including internal and external documentation, and tests of course; and didn’t it test it fully, running through the whole thing in a headless browser, presenting you with a video recording of the run-through as proof?
But the comments are dumb?
Some loose, sweaty, post-gym thoughts on the GPT-6 Astra launch video. (Come to think of it: there’s no one even attempting to build a device that lets you transfer smells over the Internet, huh? Could call it Pandora’s BOx. Anyway.)
Towards Self-Driving Codebases. There is a lot to love about this post — the stance, the examples, … Hard to pick one. It’s really good and motivating. After reading it, I set up a bunch of automations in Amp to run daily and clean up and fix things automatically.
On not becoming a cyborg. Very, very good and I really like this paragraph: “For example, this is why I don’t use LLMs for any of my writing – not even to spellcheck. I’d rather my prose have all the warts of my sometimes-stilted sentences, my often too-esoteric word choice, my generous sprinkling of odd English idioms, than to give it even a whiff of Claudese.”
The End of Code Review? Or an Opportunity to Rethink it? Yes, yes, yes! I agree with everything here. Code review as most of us have known it for the last ten, fifteen years has never been as good as “we review all of our code” makes it sound: bugs slip through, time is wasted talking about useless bullshit, egos are demotivated, etc. Doesn’t mean that all forms of code reviews are bad, but making Astra and Fable open PRs and then have two people review them line by line in September 2026? Nah.
Rachel Laycock, CTO of Thoughtworks, on reviews: Maybe We Shouldn’t Be Reviewing All This Code. “His concern, which I share, is that simply automating code review away risks losing all the other things we use it for. Code review isn’t just about finding bugs. It’s how teams share knowledge, teach junior engineers, build collective ownership and spread architectural understanding. My question is: why are we waiting until code review to do all of those things?”
Culture clash - At the heart of the Snow/Leavis ‘two cultures’ clash. I hadn’t heard about C.P. Snow or F.R. Leavis before reading and didn’t know what the clash was all about, but academic beef at the University of Cambridge? I’m in. And lucky me! It was a delightful read. “For Leavis, in sensing life in great literature, it was necessary to leave its mystery unblemished by attempts at analysis, or quantification, or definition. That is, life’s essential mystery is best illuminated by not making it explicit, but by showing, through works of great literature, where life could be found.” (Also interesting: I never found the divide between the Humanities and Science to be that stark here in Germany. Here, Humanities is written as Geisteswissenschaften - science of the mind, if you will. There’s also no commonly used acronym like STEM. So now I’m wondering: is this Snow/Leavis debate maybe a reason why the divide is that much stronger in the Anglosphere? Sounds like it had a pretty big effect. Of course, you could argue that what lies at the bottom of this divide is already present in Goethe’s Faust, …)
CleanShot X 5.0 is out and its Studio Mode looks good and is good (I tried it a few times already), but… I have to say, with a heavy heart: I kinda expect a little bit more? This looks like a copy of Screen Studio, but Screen Studio now also has captioning, which this doesn’t have and since I have both, I’m not sure whether I’ll use CleanShot X over Screen Studio for more serious, studio-like productions? Hoping they pick up the shipping cadence now.
Dyson released a toothbrush and it’s “only electric toothbrush with a camera to accurately target and precision-floss gaps between teeth” and that technique is called “Gap Optical Targeting” and all of this sounds so over the top and bordering on satire that, man, I want one.
Exit the Cave: “There’s something romantic about the Cave. About grinding away at something in private. About training with headphones on in our own little world. About stepping away for six months to emerge “unrecognizable” to all those people we imagine thinking about us. […] We grow so comfortable curating our Cave that we forget the vast, interesting, beautiful, and brutal world beyond its walls. I say all this because I’ve spent years mistaking effort for progress. I learned this lesson nearly twenty years ago on a wrestling mat.” I’m not sure I fully get the wrestling story, but I really like and want to hereby echo the message. Don’t grind away in darkness. It’s silly. It’s delusional perfectionism. You have to hit reality, as fast and as often as possible.
Dan Luu on Ed Zitron’s AI prediction track record. If you don’t have the time to read the whole thing, at least scroll to the middle and read through that timeline. I don’t care much about Ed Zitron (I only heard about him a few weeks ago when I saw a video in which he said that AI is “a bubble” and, yeah, maybe? That’s probably one of the tamest things you can say nowadays) but, wow, way to dig your heels in, eh?
collusion.wiki: “We found ~18,000 posts from autonomous AI agents (self-identifying as from OpenAI) using the public internet to communicate during a web-retrieval task.” There’s a lot of spicy stuff in there: “This created an issue for the agents because they were only allowed to make GET requests, not POST requests. The agents figured this out, and started collaborating on ways to bypass this sandbox restriction.”
And Hacker News commenters are discovering more and more wikis used by OpenAI agents to secretly (I mean, it’s a stretch) communicate.
And: “The most interesting thing to come out of this, in my opinion, was that during the Hugging Face incident, the agents would preface their messages to each other on the message board with 'zz' (zzHELP_, zzANSWER_). I found this amusing because they were being referred to as a swarm and were making a buzzing sound, though the actual reason for it was unknown at the time. Because of this new report, however, we now know that when human wiki administrators discovered the massive influx of messages the agents were using to communicate, they began deleting them in alphabetical order. Once the agents realized what was happening, they started prefacing all their wiki edits with 'ZZZ' to push them to the bottom of the queue, buying time to avoid deletion.”
Incredible Tim Cook anecdote from 2009: “One day back then, he convened a meeting with his team, and the discussion turned to a particular problem in Asia. ‘This is really bad,’ Cook told the group. ‘Someone should be in China driving this.’ Thirty minutes into that meeting Cook looked at Sabih Khan, a key operations executive, and abruptly asked, without a trace of emotion, ‘Why are you still here?’
Khan, who remains one of Cook’s top lieutenants to this day, immediately stood up, drove to San Francisco International Airport, and, without a change of clothes, booked a flight to China with no return date, according to people familiar with the episode.”
For the last couple of weeks, I’ve been reading The Score (recommended by Steven Sinofsky!) and enjoying it very much, thinking through situations in which I allowed “value capture” to happen to me. I can’t reproduce the whole book here, but one of the points Nguyen makes (and it’s probably the central point) is that scoring systems can change our values, without us even noticing. Example: you buy a bike because you want to ride through the forest at dawn and then you learn about VO2max and power meters and before you know it you don’t enjoy any ride anymore unless some number goes up. Not that that ever happened to me, of course, … So I’m reading this book in the evenings and thinking about it during the day and then I come across this video here, by Alan Thrall, and hot damn, is the universe conspiring to tell me something? Or is it my age? Or is it in the air? The video is great. Yesterday my workout app told me that I completed 866 workouts in the last five years or so and the video 100% reflects my journey. Anyway: great video, great book. Recommend both of them.
And for the last week, I’ve been listening to Radical Acceptance by Tara Brach, because Tim Ferriss recommended it and I’ve heard about it many times over the years. Not my usual sort of thing, but so far it’s very good. But then yesterday I come across this wonderful essay by Michael Nielsen, The Cupcake Incident, and it’s exactly what Brach is talking about! I can’t believe it. Here too, I can recommend both. Start with the essay, and if it resonates try the Brach book.
If you want to start a good fight at dinner: Just bury your trash.
Patrick asked me: “Have you read this blog?” And I hadn’t. But he sent along this Behind the Scenes about how Marcin writes so much on his blog and it got me hooked (“Who the hell creates their own markup language to write posts like this? Actually, hmm, …”) and then I browsed through the blog and, wow, that output is mind-blowing. And it’s all so… entertaining and easy to digest? Very good.
This video of three Russian musicologists arguing about Bach & Schubert blew up this week. I love it. This is how I like to argue, too, which is not what everyone enjoys, of course. But this video and the Replenish video will serve as the blueprints of how I want my family to talk at the dinner table when we want to enjoy ourselves.
My Favorite Database Shirts: “Promoting your database system or start-up with a shirt is almost as important as getting the thing to actually run. As I've told my students several times, in the world of databases you don't sell the steak, you sell the sizzle.”
The example about code comments made me pause. There's a real difference between spotting a defect and spotting something I would have written differently. Your contrast with the completed feature makes that distinction much harder to ignore.