Source: ChatGPT
Dear Friends,
One of my favorite descriptions of beautiful writing came from my dear old friend and colleague at The New Yorker, John Bennet. He was once describing John McPhee’s writing to me and he said, “You read it, you go along for a short bit, and then there’s an explosion. And another one.” It fits, and it’s wonderful to think about as you read McPhee, who, as close readers of my book know, was responsible in some ways for a beautiful twist in my career.
The AI at Work section below is sponsored by PwC.
This came to mind as I was reading Caity Weaver’s magnum opus, “I Found It: The Best Free Restaurant Bread in America.” I was reading it, side-by-side, with my 12-year-old son, James. At just about every paragraph, one of us would stop and mention something that made us laugh. It’s an essay about restaurant culture, class in America, and the history of bread. But it’s also just a joyous romp as Weaver moves from Joël Robuchon to Texas Roadhouse in her quest. Read it, and then read it again. Here’s just one excerpt to get a sense:
To ensure that I will be hungry enough to sample the totality of its breads at my 9:15 p.m. reservation, I consume nothing after a modest breakfast. This will prove to be a mistake. By afternoon, counting down the hours in my MGM hotel room ($39.20 a night before fees, a little more than 5 percent of my dinner bill), I pay more serious consideration to a can of Sour Cream & Onion Pringles—which I do not even like—than I did to the paperwork when I bought my car. I gaze, too, upon a lavender can next to the potato chips, envisioning the sugarplum delights it might enclose. Upon closer inspection, it turns out to contain a vibrator, two condoms, and personal lubricant (could this be edible as a kind of syrup?). By the time I am shown to my purple couch, I am hungry enough to eat the tablecloth.
I should also add that my son James, one of the finest young food critics in America, believes that the best free bread in America is served at Blue Hill at Stone Barns. I also recommend that you read the essay alongside this delightful Eater piece on the history of egg yolks, which includes an essential guide to decoding egg cartons. (News you can use!) And, because this newsletter has a rare overlap of literary enthusiasts and grizzled ultrarunners, I should also note that there is a fascinating debate right now about how much free bread you should eat during your next 50-miler. The consensus among everyone I train with is that more carbohydrates are better; you just have to train your gut to hold it all down. But there’s a counter argument that the only variable that affects performance is your level of blood sugar. Your body has enough carbohydrates in the muscles, no matter how many you take in during a race. You just have to convince your brain that you’ve got enough in the blood. If this is true, I’ll be much happier: It means I have to take one of those nasty gels an hour, not five.
I highly recommend this wild Patrick Radden Keefe caper about a mysterious spate of car crashes in New Orleans. It has, in retrospect, one of the most artful leads I’ve read in a long time. An issue or two back, The New Yorker also published David D. Kirkpatrick’s tell-all from a sympathetic ex-spy who was responsible for trying to slow Iran’s nuclear weapons program, who may have crossed from time to time to the dark side, and who may well have been wronged by a vindictive rival on the other side of a Gulf-states proxy war. It includes some of the best details about how the art of spying actually works today. If you have to, for example, hand off a key in a crowded souk, you’ll want to use “a ‘far recognition signal” like a copy of the Financial Times with its distinctive pink paper, and a “near recognition signal,” like a “a scarf tied around the forward strap of a purse slung over the left shoulder.” This is news that I hope you do not have to use.
I enjoyed this book excerpt in Vanity Fair about an alleged cheating scandal in chess—and the social media weaponization that made it go viral. You should read about how Madison Square Garden has been turned into a panopticon if you’re going to go to any games during the Knicks playoff run, which will ideally end at the hands of the Celtics in a few weeks. And you should absolutely read Daniel Engber's brilliant analysis of the case of Woody Brown, which is surely the saddest story of likely authorial fraud that I’ve read in a while.
I want to end with this remarkable speech from John Gardner, given in 1990, and particularly a passage in it that reminded me of my father:
There is no perfection of techniques that will substitute for the lift of spirit and heightened performance that comes from strong motivation. The world is moved by highly motivated people, by enthusiasts, by men and women who want something very much or believe very much. I'm not talking about anything as narrow as ambition. After all, ambition eventually wears out and probably should. But you can keep your zest until the day you die. If I may offer you a simple maxim, "Be interesting," Everyone wants to be interesting—but the vitalizing thing is to be interested. Keep a sense of curiosity. Discover new things. Care. Risk failure. Reach out.
The Most Interesting Things in Tech
One of the central questions in AI is whether demand can keep up with an industry now spending the same annually on data centers as the entire GDP of Poland. At least for now, we have our answer. Claude has been hit with multiple outages, perhaps because Anthropic underestimated how much computing capacity it will need to run, now that everyone is launching agent farms. Then there’s what happened with Allbirds. The company’s stock shot through the roof after announcing it was switching from selling sneakers to renting GPUs. This is the 2026 equivalent of adding “.com” to a business name.
A report in The Washington Post shows there has been a precipitous decline in students declaring for degrees in computer science. It’s almost certainly due to concerns over how AI will alter the value of those skills—and many of them are going into mechanical engineering instead. This is a rare example of young people pulling out of a field before adults in the industry start earning less.
Speaking of all these changes, I'm going to be in Chicago on June 2, talking about how AI agents can transform business with Glean. I would love to see you there, and I’d very much appreciate suggestions of smart people I should interview at the event. Please reply to this email with your recommendations. And on the topic of upcoming events, I will be hosting a book chat—open to fellow Stanford alumni—along with another Stanford ‘97, Tara Narula. We’ll have a conversation about our respective books, The Running Ground and Tara’s The Healing Power of Resilience. Here is the link to register.
This section is sponsored by PwC. Cut through AI noise with AI solutions from PwC and turn ideas into real business outcomes. Visit here to learn more.
To me, the last few months in AI have been the most exciting, as well as unsettling, that I’ve lived through. I’ve known intellectually that something profound was coming for a long time: exponential progress always adds up, even though the compounding is sometimes hidden. In 2022, when ChatGPT launched, I could finally start to see what this might become. And now, in the spring of 2026, I can really feel what it’s like to have extraordinary intelligence anywhere you want. I’ve set up teams of agents to do all kinds of research for me, and the work they produce is like nothing I was able to do six months ago. The future is going to be amazing, but also very confusing and fraught.
And so it’s in this context that I’m excited to announce that the fourth season of my podcast, The Most Interesting Thing in AI, goes live on Wednesday, April 22. There will be ten episodes, starting with Ariel Ekblaw, who described herself as a “space architect” when we first met. We talk about building data centers in space and the science of how we might one day live up there. I then sit down with Sam Altman to talk about how deeply we actually understand these models and their implications. I dig into the existential questions of AIs with the philosopher Nick Bostrom, and I talk with Mozilla CTO Raffi Krikorian about whether open source AI is the way to give us back our rights—or to place chemical weapons at anyone's fingertips. My goal for each conversation is to learn. I want to go into the big questions about AI that I don’t understand, and I want to try to get a handle on a technology that is moving fast. Sometimes, I feel like I’m trying to catch a squirrel with my hands, but at least I’m learning more about the way it moves.
For the first time, we also recorded them on video, so you’ll be able to watch on my YouTube channel, as well as listen on Spotify and Apple. Please do give them a listen. Let me know what you think and who else you want me to talk to.
Cheers * N