Scuttlebutt

Scuttlebutt

AI foundation models have no moat past seven days, says IBM's Vice Chairman

7 takeaways: why Gary Cohn calls the AI trade a commodity business, the fiber-optic glut he says is coming again, and Chris Sununu's case against the pollsters and the prediction markets

Scuttlebutt's avatar
Scuttlebutt
Oct 30, 2026
∙ Paid

Squawk Pod (CNBC) · Oct 29, 2026 · 41m · 34m saved

With Gary Cohn, Vice Chairman of IBM and former Director of the National Economic Council · Chris Sununu, President and CEO of Airlines for America and former Governor of New Hampshire

AI foundation models have no moat past seven days, says IBM's Vice Chairman

Photo: The White House from Washington, DC, Public domain via Wikimedia Commons

Pressed on fellow investor Steve Eisman's worry that roughly 65% of the market is now riding the AI trade on companies with no real moat, Gary Cohn didn't push back. He agreed — and said it applies to the frontier labs themselves.

"To me, they look like they're in a commodity business. No one really has a competitive advantage for more than, I don't know, seven days. Your model can be better than mine for seven days." — Gary Cohn, on Anthropic and OpenAI's foundation models

Gary, IBM's vice chairman and a former Trump National Economic Council director, laid out why he thinks the AI buildout ends the way the late-1990s fiber-optic boom did, tied it to the US government's own $40 trillion debt load, and weighed in on the Fed's next move. Later in the episode, former New Hampshire Governor Chris Sununu — now heading the airline trade group Airlines for America — made the case that pollsters and prediction markets are both reading the midterms wrong.

The full interview is covered here so you can skip it. 41 minutes of audio, 7 minutes of reading.

Here are the 7 takeaways that matter.


This Scuttlebutt Summary is available in full to premium subscribers. If you're not one already, please consider subscribing to keep reading.


This Scuttlebutt Summary is available in full to premium subscribers. If you're not one already, please consider subscribing to keep reading.


Key Takeaways

  • AI foundation models are "a commodity business" with no moat past seven days, Gary argues

  • More data centers mean more competition, not less — the opposite of what the buildout narrative assumes

  • Gary compares the AI buildout to the late-1990s fiber-optic glut: the original builders went bankrupt, second owners got rich

  • The US is $40 trillion in debt, and pension funds hold AI-era corporate debt "whether they understand they have exposure or not"

  • Gary expects a December Fed rate hike despite the pause, because "time is on the Fed's side" on inflation

  • Chris expects Republicans to end with 50 or 51 Senate seats, and says pollsters are still "polling like it's 2015"

  • Chris calls prediction markets "actually worse than the polls" because bettors just echo the media

1. The Fed's December Lean

CNBC's Steve Liesman opened the episode recapping the FOMC holding rates at 3.75%-4%, with new Fed Chair Kevin Warsh pausing after hot inflation data. Futures markets priced an 85% chance of a December hike. Gary, for his part, expects the data to force the Fed's hand regardless of what Warsh signals.

Gary expects a December hike because disinflation is still months away

"I believe it's data dependent, but I think the data will lead to a hike." — Gary Cohn, on a December Fed move

"If I'm Kevin Warsh, I'm playing for time because every day the data changes my mind." — Gary Cohn

"If you roll the clock forward into the end of the first quarter next year, oil inflation would be zero... time does help with this problem." — Gary Cohn, on why he thinks disinflation is coming

2. AI's Commodity Problem

User's avatar

Continue reading this post for free, courtesy of Scuttlebutt.

Or purchase a paid subscription.
© 2026 The Golden Age LLC · Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for great culture