Sep 9, 2026 · 1h 2m
AI’s growth promise could keep interest rates higher
GM107: What If AI Actually Pushes Interest Rates Higher? ft. Matt Klein
The episode connects uncertain AI productivity gains to bond yields, fiscal pressure, trade imbalances, and the Federal Reserve’s policy choices.
- 1AI has yet to produce a clear productivity surge, and its economic effects may take years to appear.
- 2Stronger growth, inflation expectations, and investment demand can raise long-term bond yields rather than lower them.
- 3Trade imbalances reflect domestic spending and production choices, making tariffs an imperfect remedy for global deficits.
Don't miss
Matthew Klein explains why a genuine AI productivity boom could push long-term interest rates higher rather than lower.
The brief
Matthew Klein and Alan Dunne begin with the Federal Reserve’s direction under Kevin Warsh, then ask whether AI can revive productivity as dramatically as computers and the internet did.
Klein argues that current data show no meaningful AI-driven acceleration yet, while pandemic-era hiring distortions make recent productivity weakness harder to interpret.
The central tension is that successful AI investment could strengthen growth and raise inflation expectations, increasing the returns investors demand from long-term bonds.
Klein treats U.S. debt sustainability as a question of fiscal dynamics, not an automatic crisis, and explains why trade imbalances cannot be fixed by tariffs alone.
The discussion closes on yen intervention and Warsh’s Fed task forces, especially the unresolved questions surrounding the central bank’s balance sheet.
What was said on this episode
19 statements · 8 positive · 8 negative · 3 neutral
Current data show no meaningful AI-driven productivity acceleration
“so far, we've not actually seen any meaningful productivity acceleration in the data so far”
Listen at 5:18
Productivity may accelerate in the future
“Productivity might very well accelerate”
Listen at 6:07
Business investment has increased substantially
“there has been a massive increase in business investment”
Listen at 6:18
Business investment has dramatically raised data-center input prices
“that has pushed up the prices of certain goods that are used as inputs for data centers dramatically”
Listen at 6:23
AI frontier-company wealth has affected San Francisco housing
“the housing market has definitely had a result of the spillover from the wealth boom of the AI frontier companies”
Listen at 7:00
Real interest rates rose after the 1990s productivity boom began
“real interest rates went up”
Listen at 7:41
Accelerating productivity should increase investment demand
“the demand for investment would rise relative to everything else”
Listen at 8:17
Higher investment demand should raise the cost of capital
“that should make the cost of capital go up”
Listen at 8:22
The 1990s productivity boom mainly increased growth rather than reducing inflation
“The productivity boom in the 90s mostly manifested as more growth. rather than less inflation.”
Listen at 8:55
Current conditions may be more inflationary than the 1990s
“there are a lot of reasons to think it's the opposite”
Listen at 9:36
An AI productivity repeat would not necessarily lower interest rates
“the implication that interest rates would go down, I think is not correct”
Listen at 10:15
An AI productivity boom could raise interest rates
“I think it's very plausible that if that were to happen, interest rates would go up.”
Listen at 10:18
New technologies usually take a long time to appear in productivity statistics
“the diffusion from the invention of a new technology to the actual manifestation of that, the productivity statistics, usually takes a long time.”
Listen at 11:43
Electricity took decades to produce broad process-efficiency gains
“it took decades before people understood all the ways they could update their existing processes to be more efficient”
Listen at 12:31
Users have identified many effective AI applications in software coding
“for software coding it's very clear i think people figured out a lot of the ways to make it work for them”
Listen at 13:36
AI’s economy-wide productivity gains remain difficult to determine
“the extent to which you can boost economy productivity that's harder”
Listen at 13:46
Economy-wide productivity cannot be directly measured
“there's no way of doing that”
Listen at 14:05
Pandemic employment composition made productivity appear to surge without a true boom
“doing the sort of naive math of productivity, it looked as if productivity just exploded. We had this great productivity boom. That is not what happened, right?”
Listen at 14:57
Short-term productivity movements are difficult to interpret
“I think it's very difficult to say over any kind of short-term horizon what's going on in productivity.”
Listen at 15:13
Statements are attributed to the speaker as said on the episode and reflect their view at the time, not PodLume's. They are not advice.
Books & mentions
Some links are affiliate links — PodLume may earn a commission if you buy.
