Stacy Rasgon's position going into this segment, as the anchor summarized his note, is that AI spending will not change materially after the slowdown campaign because the targets are already locked in.
Most of the debate around Dario Amodei's call to pace the frontier has been read as a safety argument. Rasgon reads the same document as a China argument, and thinks the part the US labs actually want is the part nobody is leading with.
"I actually wonder if that's one of the driving forces behind this. Frankly."
Rasgon is the senior US analyst at Bernstein, which puts him on the side of this question where a change in capex plans shows up as a change in orders.
The full segment is covered here so you can skip it.
Here are the 4 calls that matter.
👤 Guest: Stacy Rasgon, Senior US Analyst at Bernstein
🎙️ Host: Kelly Evans, who anchors The Exchange on CNBC
📰 Published: 14 September 2026 on CNBC's The Exchange
🟣 Apple Podcasts | 🔗 Episode page | ⏱️ 4 min
Key Takeaways
Spending on AI is not expected to change materially, because the targets were set before any of this
His suspicion is that limiting Chinese distillation, not safety, is a primary driver of the slowdown push
He says Amodei's essay asks for export controls to be maintained and for roadblocks against distillation
Slowing US development is not obviously the way to hold the gap with China, and he says so plainly
What is clearly on the table is managing distillation, which he thinks most US labs would welcome
Semiconductor capital-equipment sanctions have worked, on his read, and forced China onto substandard processes
The side effect is that it has made Chinese labs creative about getting results another way
Rare earths are not rare, and he treats the chokehold as solvable rather than structural
What is actually being called for right now is independent evaluators, not a stop
The pace change he describes is from extremely fast to somewhat fast on the training side
1. Capex Targets Are Locked In
Kelly Evans brought Rasgon in off the back of the previous interview, summarizing his position as "Who says don't expect spending on AI to actually change materially after this? Dario Amodei note, he says those targets are already locked in." She then put the harder version to him: what if the US decides it can hold its lead against China by slowing down, and that materially affects capex?
His first answer was about Washington rather than the labs: "I mean did you see Trump's response to this today. It doesn't sound like he's getting ready to for the U.S. to slow anything down."
Evans pushed the time horizon out past this administration: "We should. It's not crazy to look ahead two years. We don't know who the next administration will be. We don't know who the next boss will be."
The exchange set up the rest of the segment: Rasgon does not argue that a slowdown would be harmless to spending, he argues that a slowdown is not what is being asked for
2. Distillation Is the Real Aim
Rasgon's substantive answer moves the question from how fast US labs train to how much of that training leaks.
"I don't disagree that distillation is a problem. And actually, I've personally been wondering if that's one of the primary drivers of this," he said
He cited the previous guest as well as his own reading: "There is, as your prior guest said, plenty of evidence going on that the Chinese actually are getting the results they're doing by distillation, by distilling hours."
He pointed at the text rather than the headlines: "And if you actually read Dario's essay, he even talked about this on TV over the weekend as well."
The essay's China section, in his account, contains two asks: "So one was maintaining the export controls, but the other was trying to limit the amount of distillation and everything that's going, putting roadblocks in place to stop that."
His own conclusion separates the two ideas that have been traveling together: "So I don't know if slowing down the U.S. is the answer to maintaining the gap, but clearly finding some way to manage and limit distillation is clearly on the table."
On who benefits: "I think that's something that most of the U.S. labs would, would want and would welcome that. I actually wonder if that's one of the driving forces behind this. Frankly."
He did not claim to know how it would be done. Evans said it would be reasonable if there were a way; he replied "I don't know how you do it," and when she agreed, added "That's above my pay grade."
3. Rare Earths Are Not Rare
Evans took his admission badly and turned it into the broader pattern she distrusts: every time the US cracks down on China, China threatens rare earths and the US relents. "It's like it's been years of this and there's still no solution to that problem. They just go, no, no more rare earths. And we capitulate. So I don't expect too much on that front," she said.
Rasgon disagreed on the facts of that one: "Rare earths is a whole other discussion. I will say they're actually not rare. That is a solvable problem. If we want to solve it, they're there. They're dirty and polluted. But that is something that's probably solvable in the long term."
He then made the case that US restrictions have worked where they were aimed properly, pointing at semiconductor capital equipment: "Like they've been hamstrung by their ability to get competitive silicon mostly because of the semiconductor capital equipment sanctions that have put in that, that require them or forced them to make chips locally on substandard processes."
The caveat he attached is the one that leads back to distillation: "Those have as have worked very well, but it does force them to be creative and find other ways to do things which they've clearly been doing."
4. Extremely Fast to Somewhat
Evans made her own position explicit before her last question — that she likes the tools, believes the companies want to win and will, and does not trust what the doomsday talk about slowing down is trying to achieve.
Her objection is to the vagueness rather than to regulation: "If we can do things like put safeguards around practices within the industry, great. But this broad idea of slowing down like that sounds very, very vague to me. When we could have some very perhaps specific and technical solutions."
Rasgon's reply narrows what is actually on the table: "Yeah. And to be fair, the only thing that really looks to be being called for right now is this sort of concept of independent evaluators, which I don't think is a bad thing, right?"
And he pushed back on the word slowdown itself: "This whole idea of slowing down, by the way, they're not really saying like, we're not going to still be moving very fast. I think he said, we're going to move from extremely fast to moving somewhat fast on the training side."
He added that there is more to remember from a spending standpoint
Bonus Insights
The segment was built as a direct rebuttal slot. Evans handed Rasgon the previous guest's argument — "you heard we were just talking about with that researcher. He makes some valid points" — and he answered it on its own terms rather than restating his note.
The most candid moment came from the host. Evans said she does not really trust what the leaders of these companies are trying to do, immediately after saying she loves using the tools, and separated that from their ability to innovate.
Rasgon was willing to say where his expertise ends. On how anyone would police distillation in practice: "That's above my pay grade."
Evans opened the segment on a data-center timeline being pushed back and her own skepticism about whether it will really happen, which frames the capex question the rest of the interview is about.
Rasgon's bottom line is that AI spending plans are not about to move, and that the slowdown campaign is better understood as an attempt to close the distillation channel to China than as a decision to build more slowly.
Products, Companies & Tools Mentioned
Bernstein (Rasgon's firm; his note ahead of the segment argued AI spending targets are already locked in)
Anthropic (Dario Amodei's company, whose essay he says asks for export controls to be maintained and for roadblocks against distillation)
Books & Resources Mentioned
Dario Amodei's essay on pacing the frontier (Rasgon's source for the China section of the argument; he notes Amodei made the same points on television over the weekend)
If this was worth your time, send it to someone closer to the industry than you are.
Get the latest market chatter as it happens: