/AI1d ago

Cambridge researcher Herbie Bradley argues the economy will quickly adapt to AI, but founder Kanjun warns of systemic instability

Bradley cites NBER paper w34639 to support economic adaptability

120043
Original post
Herbie Bradley@herbiebradley#1012inAI

I agree it will be much faster, but on the other hand society & the economy are surprisingly adaptable to rapid change—and the overall OODA loop of the world is much faster in general than the industrial revolution

Coupled with various reasons to suspect economic impacts will be spread out over time even when capabilities exist (eg https://www.nber.org/system/files/working_papers/w34639/w34639.pdf), and the staggering of different capabilities across time, I broadly suspect there will be enough slack in the system to absorb AI's forceful change of the labor market without a huge destablilization.

I generally agree with Alex Imas’s take, but the *speed* of impact to labor is the primary difference between this wave vs the Industrial Revolution. On net it might have similar-scale multiples of economic impact, but I expect impact per unit time per worker to be much higher, which I suspect is more likely to destabilize the economic/social/political fabric.

11:37 PM · Jun 6, 2026 · 32 Views
Sentiment
Sentiment building, check back later.
Cluster Engagement
Posts from X
Most Activity
Most Activity
VIEWS11LIKES1

Interesting, I haven’t seen this NBER paper. I think I could make an argument either way (for destabilization vs adaptation to rapid change).

However, I’m less optimistic about adaptation to rapid change because of 1) low state capacity at least in the US and 2) market drives for perpetual corporate growth without regulatory intervention, which in any biological system would be considered a cancer.

AI + internet enables a much higher rate of wealth accumulation, and I suspect will lead to much increased income inequality. This tends toward a less stable society unless part of the population is oppressed.

Herbie Bradley@herbiebradley

I agree it will be much faster, but on the other hand society & the economy are surprisingly adaptable to rapid change—and the overall OODA loop of the world is much faster in general than the industrial revolution

Coupled with various reasons to suspect economic impacts will be spread out over time even when capabilities exist (eg https://www.nber.org/system/files/working_papers/w34639/w34639.pdf), and the staggering of different capabilities across time, I broadly suspect there will be enough slack in the system to absorb AI's forceful change of the labor market without a huge destablilization.

9hViews 11Likes 1Bookmarks 0