Substack author. Plain English podcast host. Abundance co-author. Contributing writer at The Atlantic.
Washington, D.C.
- Tweets 23,072
- Following 2,136
- Followers 264,862
- Likes 38,500
Pinned Tweet
Some personal news.
Today, I’m leaving The Atlantic after almost 17 years and moving my writing to Substack.
It would be convenient, for the purposes of crafting an exciting departure announcement, to have a dramatic exit story: a fight, a grievance, a shouting match with an editor that ended with me hurling a bunch of leather-backed Thoreau volumes across the open-plan office. That is not the case here.
I love The Atlantic, and I'll remain a contributing writer there. But after almost two decades at one publication, I wanted to write for myself. The things I've published that I'm most proud of—whether it was the original abundance agenda essay, or my piece on workism—emerged from a very personal expression of frustration, or confusion, or curiosity. I want to know what my thinking and writing is like if I lean into a more independent and personal writing life.
That's brought me to Substack, which is already home to an astonishing share of my overall reading. I'm excited to join their community and excited to build my own. The name of the newsletter should be easy to remember: Derek Thompson.
The newsletter will have three main pillars
1. Abundance
2. The frontier of science and technology—GLP1s, AI, biotech, energy breakthroughs—covered in a way that’s both curious and skeptical
3. The anti-social century & the social crises of anxiety and aloneness
Thanks to The Atlantic for 16.8 incredible years and thanks to everybody who follows me across the river.
- dt
325
265
6,677
1,656,644
Derek Thompson retweeted
I'm thrilled to be launching Econvitals today. It's been the main way in which I have been tracking and understanding the economy over the last several months. And I decided I might as well share it with all of you--for free.
Has a combo of website, email lists & social media.
17
97
506
74,964
Derek Thompson retweeted
The best part of my job is working with writers I've admired for years. Thank you @DKThomp for having me!
Listen for our conversation about cracks in the AI thesis.
New pod/newsletter: THE AI QUESTION THAT THE VALLEY FEARS AND WALL STREET CAN'T ANSWER:
How much external revenue is the AI ecosystem actually generating?
“All I care about is are the revenues coming into the system greater than the costs of building the whole value chain,” tech analyst Craig Moffett recently told @benthompson. “The market is panicking right now, because [it's] saying, ‘We don’t know.’"
I talked to Ramp economist @arakharazian about what his firm, which oversees billions in AI corporate spending, is seeing. He's worried, too. Kharazian sees several "cracks" in the AI thesis:
1. AI spending is highly concentrated among VC-backed tech companies
The top 1 percent of AI spenders accounts for 80 percent of all AI spending. They are highly likely to be backed by venture capital. Why that's awfully interesting: AI demand is supposed to account for revenue generated outside of the AI capital cycle. But if most AI spending comes from companies that rely on venture capital firms that also rely on AI valuations, this seems like another example of the AI capital cycle paying itself.
2. Spending among the top 1% of companies is stagnating
Per-employee spending among Ramp’s top 1 percent has fallen in the last few months. Token spending—that is, the money businesses pay to run models through their software—recorded “two straight months of virtually no growth,” Kharazian wrote.
3. Falling token prices
Intelligence getting cheaper is good for many parts of the AI ecosystem. But it's a challenge for external revenue, as firms get better at avoiding the most advance models to find cheaper tools to do most of their work.
piped.video/watch?v=e7F1AvDs…
2
5
34
10,052
New pod/newsletter: THE AI QUESTION THAT THE VALLEY FEARS AND WALL STREET CAN'T ANSWER:
How much external revenue is the AI ecosystem actually generating?
“All I care about is are the revenues coming into the system greater than the costs of building the whole value chain,” tech analyst Craig Moffett recently told @benthompson. “The market is panicking right now, because [it's] saying, ‘We don’t know.’"
I talked to Ramp economist @arakharazian about what his firm, which oversees billions in AI corporate spending, is seeing. He's worried, too. Kharazian sees several "cracks" in the AI thesis:
1. AI spending is highly concentrated among VC-backed tech companies
The top 1 percent of AI spenders accounts for 80 percent of all AI spending. They are highly likely to be backed by venture capital. Why that's awfully interesting: AI demand is supposed to account for revenue generated outside of the AI capital cycle. But if most AI spending comes from companies that rely on venture capital firms that also rely on AI valuations, this seems like another example of the AI capital cycle paying itself.
2. Spending among the top 1% of companies is stagnating
Per-employee spending among Ramp’s top 1 percent has fallen in the last few months. Token spending—that is, the money businesses pay to run models through their software—recorded “two straight months of virtually no growth,” Kharazian wrote.
3. Falling token prices
Intelligence getting cheaper is good for many parts of the AI ecosystem. But it's a challenge for external revenue, as firms get better at avoiding the most advance models to find cheaper tools to do most of their work.
piped.video/watch?v=e7F1AvDs…
14
18
116
23,347
A concluding thought:
In 2020 and 2021, investors looked at Peloton’s and Zoom’s explosive growth, projected their revenue curves into the future, and concluded that these companies would take over fitness and work. Instead, what the pandemic really did was concentrate several years of demand into a few quarters. The earnings that followed disappointed, and their stocks fell.
I'm not an AI bear. But I wonder if the explosion of coding agents did something a little bit like this for AI. Software companies adopted Claude Code and Codex so quickly—and spent so lavishly in their adoption—that the most valuable AI market was saturated really fast. Annualized revenue more than tripled. But tripling revenue again might probably require a new S-curve, a coding-agent revolution for finance, law, or manufacturing.
“The models are becoming so advanced at solving math and physics problems every month that maybe there will be another coding agent revolution and there will be some other form in which AI augments a new sector,” Kharazian said. The multi-trillion-dollar AI bet might depend on it.
derekthompson.org/p/the-ai-s…
2
1
16
6,766
Kalshi’s over/under on “columns after the midterms that declare American politics to be ‘Post-Trump’” currently screaming toward 10,000.5
Axios: Senate Republicans fear Mississippi is in play
Senate Republicans are weighing three last-ditch spending moves that would have been unimaginable six months ago: a cash infusion for Mississippi and even more money for South Carolina and Kansas.
"The wall has broken. The enemy is inside the fort," one national Republican political operative told Axios. "We are now fighting hand to hand."
axios.com/2026/10/21/mississ…
3
2
149
31,803
I suppose it’s a testament to Fielder and Oppenheim that every review of You Can See Everything claims to see a different lesson or moral in the movie…
It’s about the post-truth era, or it’s maybe even about Trump or Musk, it’s about the limits of the documentary, it’s about the nature of lying, its about the psychology of psychopathy, it’s about Silicon Valley’s hype economy, it’s about … autism.
I don’t think it was principally about any of those things!
Very confused by Tyler's take here. I think the movie really undersold just how clear the case against her was, maybe to lull us in more.
23
18
434
64,184
I think what makes the movie both incredibly powerful and also frustrating to people who know the Holmes story really well is that it’s a film mostly “about” the impossible ambiguity of truly getting to the bottom of another person*—and that lesson is largely true and it’s profoundly performed in the movie—but this particular movie is about a very particular person who lied so blatantly in so many ways that an alternative and more prosecutorial documentary wouldn’t have struggled much to prove who she is: a convincing liar who’s lying (convincingly at times) about having been a convincing liar. And so while Fielder’s artifice is both incredibly entertaining and often ingenious I get the critique that it’s also an unnecessary maneuver to get to the bottom of *this particular person* whose record of lying is so blatant and whose psychopathy is so richly documented in the public record.
*I mean, right? That’s clearly the point? Seems important that the movie begins with the husband saying “nobody knows who she is” and ends with Amanda saying she can’t be sure who Elizabeth is or how Nathan feels about her, and the last frame is a paragraph of text about how no one—not the NYT, not Fielder, not Holmes—can be absolutely sure if Elizabeth actually told the Times that Amanda was “playing a character I created” bc nobody has a perfect record of the most important quote in the film ever actually being said. The movie is designed to build toward the conclusion that there is no definitive conclusion.
15
2
123
19,225
My take on wealth taxes is they won't work well at the local or state level, because you're targeting the people with the greatest means and motivation to just pick up and move somewhere, and they won't work well at the national level, because they'd almost certainly require a constitutional amendment, just as federal income taxes did (the 16th Amendment), so in both cases you're creating a levy that is basically designed to disappear.
California's Billionaire Tax Act would collect roughly $30 billion. Accounting for lost income taxes from departures, the mean Net Present Value is negative $39 billion, from Benjamin Jaros and Joshua Rauh nber.org/papers/w35899
119
72
1,074
178,942
Lot of people furious at this tweet but the amount of energy required to pass a national wealth tax that the Supreme Court will just overturn in 6 months is much much much better spent repealing the Trump tax cuts, creating a higher minimum effective tax rate for the wealthy, doing cap gains, or trying payroll tax ceiling increases.
We have like infinity legal and proven ways to actually raise revenue for welfare and deficit reduction, I don’t understand the enthusiasm for rejecting these in favor of something we know won’t work
12
9
175
15,987
Derek Thompson retweeted
Fascinating podcast. Try to make sure my AI listening diet is not just doomer-stuff, and there was something nice listening to two guys geeking out over the technology.
Not sure I'm fully convinced about the societial implications, but lots of food for thought.
New pod: 'YOU'RE NOT AGI-PILLED ENOUGH'
I asked @emollick to tell me what I'm getting wrong about AI. We also talked about:
- why so many experts keep underrating the rate of progress for model capabilities
- why so many experts keep *overrating* how rapidly powerful AI will change the macroeconomy
- why the answer to the previous question starts with "O-rings and garbage cans"
- the sociology of AI swarms
- the paradoxes of AI use (or: 'every technological augmentation is also an amputation,' per McLuhan)
- the end of prompting (or: what happens if personal AI agents get so good at anticipating our needs that humans, once the prompter, become the prompted)
- 2027 predictions and ... his excellent new book CO-EXISTENCE
piped.video/watch?v=zthE5Oa8…
1
8
9,023
Whether AI is crowding out the physical world economy, or just the physical economy's bright shiny interest-rate-insensitive diamond in the rough, or somewhere in between, the situation is that "homes for chips" spending is on fire and "homes for people" spending is putrid
11
15
166
26,021
Derek Thompson retweeted
What do Larry Kudlow, Janet Yellen, Steve Mnuchin, @econJaredB & me have in common? We all believe the permitting reform bill is essential for America to build, grow, tackle climate change and maintain our technological lead.
Probably the broadest letter I've ever signed.
15
59
286
38,954
Derek Thompson retweeted
This was a great conversation!
New pod: 'YOU'RE NOT AGI-PILLED ENOUGH'
I asked @emollick to tell me what I'm getting wrong about AI. We also talked about:
- why so many experts keep underrating the rate of progress for model capabilities
- why so many experts keep *overrating* how rapidly powerful AI will change the macroeconomy
- why the answer to the previous question starts with "O-rings and garbage cans"
- the sociology of AI swarms
- the paradoxes of AI use (or: 'every technological augmentation is also an amputation,' per McLuhan)
- the end of prompting (or: what happens if personal AI agents get so good at anticipating our needs that humans, once the prompter, become the prompted)
- 2027 predictions and ... his excellent new book CO-EXISTENCE
piped.video/watch?v=zthE5Oa8…
11
7
185
32,299
Derek Thompson retweeted
What else distinguishes the 1% from everyone else?
Having a stay at home spouse...
What distinguishes the 1% from the merely affluent? How about having a third child.
New post on a topic I've long been fascinated by (the economics and financing of having a third kid) where I finally felt bold enough to go into the latest SCF microdata to get a good answer. /1
44
76
1,752
201,077
New pod: 'YOU'RE NOT AGI-PILLED ENOUGH'
I asked @emollick to tell me what I'm getting wrong about AI. We also talked about:
- why so many experts keep underrating the rate of progress for model capabilities
- why so many experts keep *overrating* how rapidly powerful AI will change the macroeconomy
- why the answer to the previous question starts with "O-rings and garbage cans"
- the sociology of AI swarms
- the paradoxes of AI use (or: 'every technological augmentation is also an amputation,' per McLuhan)
- the end of prompting (or: what happens if personal AI agents get so good at anticipating our needs that humans, once the prompter, become the prompted)
- 2027 predictions and ... his excellent new book CO-EXISTENCE
piped.video/watch?v=zthE5Oa8…
8
18
199
95,782
Full transcript: derekthompson.org/p/everythi…
6
9,100