Siri tries to predict the next word

This week's pod is with the great @kait.bsky.social. It's about what phones/screens are really doing to us (and what we can and cannot prove). We talk about ditching smartphones for flip phones, all the debates, phone bans, social media lawsuits. And why simple answers are hard to come by.

Two of my favorite people discuss doing phone. Listen up!! www.theatlantic.com/podcasts/202...

Statement from The Atlantic's editor in chief Jeffrey Goldberg:

I dunno that I've had a single conversation this year with anyone about media and not cited the work and philosophy of @pablo.show as an example of how to break the mold and do something new and meaningful. One of my fav conversations talking about it with him on the pod this year

In OpenAI's blog for the GPT-5.5 Instant model, an example is tea recommendations in San Francisco. They are odd. The person likes tea house over boba. But the top rec is Cere Tea which is basically matcha lattes, einspanners, etc—sugary boba, not tea house. Also don't think Cere serves oolong?

The soft nationalization continues, @lilashroff.bsky.social and I wrote about how this all plays out recently: www.theatlantic.com/technology/2...

A little late to posting this but really really good and important www.thecut.com/article/wego...

After three days of trial pretty clear Elon doesn’t know the basics of AI safety (testifying under oath he’s unfamiliar with the terms “preparedness framework,” “CAISI,” etc) and is indifferent to how xAI poses the same risks to the world his lawsuit alleges OpenAI does (basically testified this)

This really is amazing: www.theatlantic.com/health/2026/...

Day one of Musk v. Altman, a drama-filled legal war over a feud that has shaped the entire AI boom: www.theatlantic.com/technology/2...

Anthropic used to play the role of younger brother to OpenAI—but now Sam Altman is remaking his company to look more like its rival, @matteowong.bsky.social reports:

In @nytopinion.nytimes.com America’s electric grid is old and could soon be pushed past its breaking point. “Something needs to change now,” Robinson Meyer, the founding executive editor of Heatmap, writes.

I understand it's ramp season but "better than pesto"?

Anthropic and OpenAI could challenge the federal government’s monopoly over power and unsettle its ability to provide for its citizens. What happens a if the White House has enough and tries to commandeer the top AI firms? @lilashroff.bsky.social and I explore:

Loveee Lousia Thomas on chess (But are we "After Carlsen?" He just won a major speed chess tournament playing... on a tablet (a major handicap)) www.newyorker.com/sports/sport...

Brilliant horrifying episode—but if you get sad like I did listen to the end for Charlie’s rallying cry for humanity

huh

"Ironically, America doesn’t seem to need much help when it comes to disappearing scientists." www.theatlantic.com/science/2026...

I say this a lot but today’s pod is worth your time. It’s with Josh Owens who worked for Alex Jones/Infowars for 4 years in the 2010s. It’s a conversation about how the conspiracy machine works from inside, about radicalization and, crucially, about the hard, unscalable work of deradicalization

MAJOR UPDATE: I found the best free restaurant bread in the United States www.theatlantic.com/magazine/202...

this is gold (and a fascinating technical explanation of the debt financing here) www.bloomberg.com/opinion/news...

What happens when agents can do schoolwork? My latest for @theatlantic.com on the next phase of AI in the classroom: www.theatlantic.com/technology/2...

Claude Mythos Preview is, of course, a big deal for cybersecurity. But perhaps more importantly, the model brings into stark relief how AI companies are amassing more power and influence than nation-states. www.theatlantic.com/technology/2...

I suppose we’re finally about to find out www.theatlantic.com/technology/a...

something strange and horrible and somehow fitting that we should have a very real threat of madman civilizational destruction at the very moment when we also have humans on the dark side of the moon taking pictures that show how small, precious, and beautiful our world and existence is

I needed the world to know that Polymarket has an X account for its female traders called @PolyBaddies. My latest for @theatlantic.com : www.theatlantic.com/technology/2...

Here comes RAMageddon!! Very good analysis here from @hana-kiros.bsky.social www.theatlantic.com/technology/2...

On AlphaGo, reasoning models, automation, life: www.theatlantic.com/technology/2...

I have felt a little nuts listening to experts and then seeing how, broadly, the world hasn't caught up to the severity of all this. but it really seems like when the gap begins to close, it is gonna close fast. something @matteowong.bsky.social and i tried to get at here:

sharp and essential reading

we've made over 20 of these pods as of this week...it has been a fun/humbling/chaotic learning experience trying a different medium. anyhow! if you have been enjoying this at all or have thoughts about it, my dms are open! but also...maybe drop us a like/subscribe/idk one of those things. it helps!

Omg??????? www.nytimes.com/2026/03/26/b...

The AI economy looks...really precarious. So @matteowong.bsky.social & I did a bunch of reporting to try to figure out what happens when a potential bubble collides with a war in Iran and a potential resource shortage. The answer is...arguably the most dire stuff i've heard from smart ppl in a while

The AI boom's finances had already become a whole lot more precarious over the past six months. Enter a war with Iran that threatens several of the key industrial inputs to the shockingly fragile data-center supply chain. With the brilliant @cwarzel.bsky.social: www.theatlantic.com/technology/2...

"... In time, those Unconscionable Maps no longer satisfied, and the Cartographers Guilds struck a Map of the Empire whose size was that of the Empire, and which coincided point for point with it ..." www.bloomberg.com/opinion/news...

correct www.theatlantic.com/culture/2026...

That’s right and it’s the best. (But cube your potatoes don’t cut them in rounds, that’s weird and small chunks are the ideal textural and fork experience!!)

I wrote about OpenAI’s ongoing identity crisis: www.theatlantic.com/technology/2...

"It’s because the U.S. Constitution has failed that Claude has a constitution" www.newyorker.com/magazine/202...

Lovely little piece by @matteowong.bsky.social about Anthropic's sponsorship of a Claude Monet exhibit, and the greater meaning of "interaction" that the company (and its technology) misses

"I thought they'd be a nice durable cardboard," the owl-eyed man says of Gatsby's library. Anyway, Anthropic made a Claude AI typewriter to go alongside an exhibition about Claude Monet. I tried it. www.theatlantic.com/technology/2...

Our story on Meta's legal woes by @kait.bsky.social

www.newyorker.com/magazine/202... “We won’t know the harm fully until it’s too late.”

AI services are kind of weird. They're not like web browsers, operating systems, social networks, or even smartphones. They are becoming more like commodities—facial tissues, catered slop-bowl trays, or electricity. I wrote something about this:

"Sam Altman has predicted that large language models will soon be capable of 'fixing the climate, establishing a space colony, and the discovery of all of physics" — but, "might be able to extrude only something equivalent to 'a real poet’s okay poem'" @jasmine.bsky.social on what sets writing apart