Top of the Morning: The AI Economy 4Q26 update - Part 2

Andrew Dubinsky · UBS

AI isn't yet visibly disrupting aggregate US jobs; the damage is concentrated in high-tech and young workers.

UBS's view is that AI is not yet visibly disrupting the aggregate labor market. The unemployment rate has drifted lower since the spring AI-layoff headlines, and occupations and industries with high AI exposure are not doing worse in aggregate than low-exposure ones. The pressure shows up in pockets. BLS data show high-tech employment (software developers, computer systems designers) down about 1.4% year over year, versus modestly positive job growth overall. Within highly exposed occupations, workers in their early 20s and 30s show mid-to-low single-digit job contraction year over year, per Stanford Digital Economy Lab data. Construction employment in AI-exposed areas, by contrast, is growing about 3% year over year and accelerating, which mutes the aggregate impact.

“we just weren't seeing huge signs of it in the aggregate job growth and really had to zoom in on certain sectors.”at 1:15

Andrew Dubinsky · UBS

AI stock gains are lifting US consumer spending through the wealth effect by a few tenths of growth.

AI-linked stocks have risen sharply, and this has made households feel wealthier. That wealth effect drives more spending. Andrew estimates that at least a few tenths of extra consumption growth can be linked to AI stock gains. This is the indirect channel through which AI feeds US growth, alongside the direct investment spending he describes.

“that wealth effect leading to more spending. So we're looking at at least a few tenths of extra consumption growth that are, I would say, are linked to AI stock gains.”at 4:29

Andrew Dubinsky · UBS

AI-related investment is a resilient tailwind, adding about three quarters of a point to US growth.

AI affects US growth through two channels. The first is indirect: AI stock gains produce a wealth effect that lifts consumption. The second is direct: AI-related equipment and other investment appear in the national accounts. After adjusting for inflation and for the import content of much of this investment, that spending has added about half a percentage point to GDP growth over the past four quarters. Combined with the wealth effect, AI adds roughly three quarters of a point to growth. Andrew describes this as a resilient tailwind and one reason the US has absorbed energy shocks smoothly.

“...that equipment spending ... adds up to around half a percentage point after you adjust for the fact that much of the investment involves imported goods... So together, those two effects are adding up to, let's say, three quarters of a percentage point of growth and a pretty resilient tailwind.”at 4:59

Andrew Dubinsky · UBS

AI demand is still net inflationary through elevated digital goods prices, a key concern for the Fed.

Digital goods prices, which used to deflate at a steady pace, are still rising, and part of that is AI demand. Some of the recent price movement was revised away, but the underlying trend shows positive inflation. Andrew expects more supply to come online, easing key bottlenecks and slowing digital goods inflation. He flags an open question about whether AI ultimately becomes a source of disinflation, deflation or inflation. For now he judges AI to be net inflationary. The Fed is watching this as it weighs tightening policy, and the AI demand story has not faded yet.

“For now, it's still net inflationary. And one of the key themes the Fed is watching when it's thinking about tightening policy.”at 6:24