Moltbook is weird and fascinating and scary. It's also a signal about the present-day internet as much, if not more than, a glimpse into its future

"One student came up to me after the first flash essay, a little frustrated and worried that he wasn’t meeting the mark. He felt like he was writing into the complete unknown, rather than with a plan in mind. I said that was exactly the point." www.theatlantic.com/ideas/2026/0...

From this morning: www.theatlantic.com/technology/2...

The end of the AGI consensus: www.theatlantic.com/technology/2...

Artificial general intelligence was supposed to have been imminent—but “the broad agreement on what AGI even is and the immediate value it could provide humanity have been scrubbed away,” @matteowong.bsky.social reports:

What does Anthropic really stand for? The AI company “comes across as more sincerely committed to safety than its competitors, but it is also moving full speed toward building tools it acknowledges could be horrifically dangerous,” @matteowong.bsky.social reports:

Anthropic wants to be the AI industry's superego, but is caught between the pressures to be safe and fast, rigorous while being commercially successful. I profiled the company and its leadership, who seem earnest but torn, anxious but at times hubristic: www.theatlantic.com/technology/2...

Given the grok horrors of the past few weeks and the crisis of impunity we’ve got on our hands, today’s pod is with the excellent @sophiegilbert.bsky.social on the ways that misogyny has been coded into these platforms and shaped our culture

2025 saw the most AI child sex abuse in history—and likely the least amount of it for years to come:

look at these responses for yourself and see that few people can even be bothered to even respond (one investor rep seemed confused about whether they had actually invested?). they are banking on everyone moving on. cowardice as a crisis comms strategy.

There is a crisis of impunity. The Musk/Grok stuff is just one f many examples of people who are counting on there never being any consequences. And it all makes me just incandescently mad.

Musk and xAI enabled widespread sexual harassment on X. @cwarzel.bsky.social and I asked xAI's key investors, as well as Google, Apple, Nvidia, and more: Do you endorse the use of Grok to undress people? If not, will you cut ties with xAI? Few replied; none answered our actual questions.

We are not letting this go If there is no red line around AI-generated sex abuse, then no line exists. by @cwarzel.bsky.social and @matteowong.bsky.social

👑👑👑

New from @garymarcus.bsky.social and me— Chatbots are: 1. Bad at providing info during breaking news events 2. Used by many people to find that info anyway 3. Making it harder for journalists to get their material seen online 4. Great vectors for manipulated media and propaganda Not great!

NEW: Trump threatened Venezuelan leader Delcy Rodríguez in a phone interview this morning with me: “If she doesn’t do what’s right, she is going to pay a very big price, probably bigger than Maduro." www.theatlantic.com/national-sec...

The moon seems to have a dedicated PR team, with news outlets hyping Wolf Moons, Super Blood Moons, Cold Moons, and more, Kaitlyn Tiffany writes. She explores what's behind the moon hype:

Really can’t get over what’s been happening on X. 1. Man asks Grok to put young teenagers in “sexy” underwear 2. Grok does it 3. I ask Grok how old the people in the photo are 4. Grok says “13-16 years old” despite having just made a “sexy” image of them 5. Man deletes post

People are using Grok to generate non consensual sexual images of women and apparent minors—and X has turned that harassment and abuse into a meme. Elon, meanwhile, is cracking jokes about it all

The MetroCard became metonymic with NYC. At midnight today it will retire, and so in MetroCard’s honor I learned about how this under appreciated technological marvel came to be and saw how MetroCards are produced. A eulogy:

Honored to work with the most brilliant journalists, who also reliably produce the most mind blowing facts!!

I had a blast with this piece, interviewing such talented jingle writers—including the man who is literally the reason Blue Öyster Cult put more cowbell on “Don’t Fear the Reaper” www.theatlantic.com/technology/2...

“God sculpted the Bay,” he pronounced. Now the Bay is sculpting God. Or at least, a bunch of other 20-somethings in San Francisco think they are. Amazing reporting and writing from @lilashroff.bsky.social: www.theatlantic.com/technology/2...

The white-supremacist influencer Nick Fuentes is more powerful than ever and is “laying groundwork to go even bigger,” Ali Breland reports. Read what Breland learned after a marathon viewing of Fuentes’s show:

Has been a tough few months for OpenAI: www.theatlantic.com/technology/2...

My feature on Jollibee is out today!!! Read on for a tale about fried chicken, sweet spaghetti, and postcolonial revenge 🇵🇭 www.theatlantic.com/magazine/202...

A new study asked whether an AI model can influence people’s political inclinations. @matteowong.bsky.social explores what the findings mean for the campaigns of the future:

"36–23 still taunts and tantalizes us, a magical missing score that reminds every Scorigami devotee that what’s possible isn’t preordained. That’s why National 36–23 Day will be such a celebration, if it ever comes." www.theatlantic.com/science/2025...

“he seemed to be suffering under the strain of his belief that the creation of smarter-than-human AI was imminent and that it would almost certainly lead to the end of all human life. He often talked about the possibility that AI could kill his sister, and he seemed to be motivated by this fear.”

Amidst more and more cases of "AI psychosis," I talked to psychiatrists to figure out we do—and more importantly don't—know about how chatbots may be producing or exacerbating severe psychological distress. www.theatlantic.com/technology/2...

"Something disastrous happened here, and academics are nearly united in the opinion that the problem is not simply a product of the pandemic." nymag.com/intelligence...

There's an all-out AI arms race afoot between hackers and defenders. In the near term at least, cybercriminals may be coming out ahead: www.theatlantic.com/technology/2...

Stellar and essential reporting on the incentives and decisions on ChatGPT www.nytimes.com/2025/11/23/t...

wrote with @matteowong.bsky.social about Elon's ritual humiliation at the hand of his chatbot, Grok. It's hilarious, yes. But it's also a reminder that he's trying to bend his platform/tools to rewrite or warp history (gift link) www.theatlantic.com/technology/2...

Bless @sophiegilbert.bsky.social @theatlantic.com: "The past decade has been a gloomy lesson in how limited a proportion of men actually see women as equal human beings...The fish rots from the head. The pig is in the Oval Office." www.theatlantic.com/culture/2025...

A sprawling criminal enterprise is responsible for all of those fake DMV texts, and it has a name: the smishing triad. @matteowong.bsky.social talks to the cybersecurity experts trying to take it down: https://theatln.tc/aQKBYgGA

Of course even those fake toll fee texts are the result of SaaS—that is, Scamming-as-a-Service www.theatlantic.com/technology/2...

"The science backing up the idea that race can make someone a good or bad scientist or airline pilot is as solid as the logic behind 'orcs can’t be wizards' or 'a hobbit can never become a great fighter.'" Just an excellent piece

I’m still fighting every day to get the answer to this question and now I have a FORM. Please for the love of God fill out the form (or email your answer to cweaver@theatlantic). I’ll let you know how it all works out in the end!!!! docs.google.com/forms/d/11de...

OpenAI requested the names of anyone who cared for or supervised Adam Raine—whose parents allege, in a lawsuit, killed himself at ChatGPT's encouragement—over the past 5 years, including friends, teachers, school-bus drivers, and “car pool divers" [sic]. www.theatlantic.com/technology/2...

OpenAI is facing several high-profile lawsuits—and now the company is going on the offensive. @matteowong.bsky.social reports on the tech giant’s new, aggressive legal strategy:

Hi! Preposterously excited (& tbh nervous in the good but also terrifying way) to say that i'm spinning up a podcast that will expand on all the work i do covering politics/technology/information hell. here's a lil teaser we made - first ep will come out 11/14! www.youtube.com/watch?v=lnDE...

In the end, Andrew Cuomo’s long record was a gift to Zohran Mamdani. “What I don’t have in experience, I make up for in integrity—and what you don’t have in integrity, you could never make up for in experience,” Mamdani told Cuomo in a debate.

Defining and essential from @damonberes.com on how AI is ushering the age of anti-social media: "Although chatbots may be built on the familiar architecture of engagement, they enable something new: They allow you to talk forever to no one other than yourself." www.theatlantic.com/magazine/202...

More essential @ibogost.com on the state of higher education—an incredibly smart, compassionate, and compelling view on the specter of "grade inflation" www.theatlantic.com/technology/2...

“We can’t police that whole thing,” Common Crawl said. “It’s not our job. We’re just a bunch of dusty bookshelves.” Meanwhile, CC has accepted hundreds of thousands in donations from AI companies such as OpenAI and Anthropic. And it expressed open antagonism toward the media:

Be sure you get to the kicker in this essential investigation into Common Crawl, the data archive behind the AI boom

Zohran Mamdani argues that “freeze the rent” is not just a way of delivering relief from exorbitant housing costs—it is the only way to get enough voters on board with a growth agenda. Rogé Karma spoke with Mamdani to see if he has a point.