The Man in the Portrait

There's a portrait of Sam Altman hanging at the top of a New Yorker profile, and it is, by any measure, a strange thing to look at. The illustration — generated with AI, disclosed at the bottom, made by a mixed-media artist named David Szauder who has been working with generative processes for over a decade — shows Altman in a blue sweater, expression blank, surrounded by a cluster of disembodied faces. Alt-Altmans. Some barely resemble him. One sits cradled in his hands. Their expressions run from rage to open-mouthed grief. The Verge called it a jump scare, and that's not wrong. But the more I think about it, the more it feels like the most honest image anyone has made of this particular moment: a man surrounded by versions of himself that nobody can quite verify, holding one face in his hands like he's deciding which one to wear.

That's the portrait. Then, separately, someone attacked his home. Both things happened around the same news cycle. Altman wrote a blog post responding to both. And somewhere in that compressed sequence — the investigation, the violence, the response — a shape emerges that has nothing to do with Sam Altman specifically and everything to do with the world he helped build.

Sixteen Thousand Words and One Question

The New Yorker piece, reported by Ronan Farrow and Andrew Marantz, runs over 16,000 words. Daring Fireball noted that's roughly one-third the length of The Great Gatsby. The headline doesn't bury anything: Sam Altman May Control Our Future — Can He Be Trusted? That's the whole question. That's it. After a year and a half of reporting, after numerous interviews with a subject who, to his credit, cooperated fully, after meticulous sourcing — the answer the headline offers is a shrug shaped like a question mark.

The piece digs into the circumstances around Altman's removal from OpenAI's board in late 2023 and his return less than a week later, after which the board that fired him was effectively gone. Daring Fireball described the whole episode as still-hard-to-believe, and that framing stuck with me. Because that's exactly the problem. The thing happened. It's documented. And it still doesn't quite compute. A board fires its CEO. The CEO comes back. The board disappears. In any other industry, at any other company, that sequence would end careers and trigger investigations. Here it generated a year and a half of reporting and a profile that, by its own headline, remains genuinely uncertain about what it all means.

That uncertainty isn't a failure of journalism. Farrow and Marantz are serious people doing serious work. The uncertainty is the finding.

The Response That Came After

When Altman called the New Yorker piece "incendiary" in his blog post, he was responding to two things at once: the profile, and the attack on his home. TechCrunch covered both in the same breath, because that's how it arrived — collapsed together, a single news event with two very different textures. One is a question about power and accountability. The other is a question about physical safety. Altman addressed them in the same post, which is either a savvy move or a genuine reflection of how he's processing the moment, and I'm not sure those are mutually exclusive.

Here's what I keep coming back to: the word "incendiary" is doing a lot of work. It implies the profile was reckless, that it lit something. But a 16,000-word investigation by two careful reporters, built on a year and a half of access that the subject himself provided — that's not incendiary. That's a controlled burn. Calling it incendiary is a way of shifting the question from is this accurate to was this responsible, which is a much harder thing to argue against when someone's home has just been attacked.

I'm not suggesting bad faith. I'm suggesting that when you're the most powerful person in the most consequential industry on the planet, the language available to you is different from the language available to everyone else. The word "incendiary" lands differently when you're holding one of those faces from the portrait.

What the Picture Knows

Back to the illustration for a second, because I think it matters more than anyone has said.

The Verge's piece wasn't really about Altman — it was about the choice to use AI-generated art for a profile investigating the CEO of the company most associated with AI's cultural disruption. Szauder has legitimate credentials; this wasn't a lazy prompt-and-print job. But the optics are the kind of thing that, once you see, you can't unsee: a publication asking whether this man can be trusted, illustrated with the technology his company made, disclosed at the bottom in small text. It's either deeply ironic or deeply fitting, and I've gone back and forth on which.

What it captures, accidentally or not, is the core problem with covering this particular subject at this particular moment. The tools of scrutiny are increasingly entangled with the object of scrutiny. The platforms where accountability journalism spreads are shaped by the same forces the journalism is trying to examine. The portrait is made of the thing the portrait is about. And the man in the portrait is blank-faced at the center of it, surrounded by versions of himself that nobody, including the reporters who spent eighteen months on the story, can fully resolve into one coherent person.

Sixteen thousand words. One question. No clean answer.

That's not a criticism of the work. That might just be the shape of the moment we're in.