The AI bubble that economists refuse to see

Goldman Sachs calculated (early 2026) that AI contributes “basically zero” to the economy. This isn’t somewhat “disappointing” or a bit “below expectations”… this is zero point zero. For anyone who doesn’t get it: this means NOTHING! And what are all those educated economists saying now, the ones who nodded along for months when someone who fancies himself some kind of god in the White House claimed AI “drives half our growth”? They’re acting like this is news. As if they didn’t know exactly that the numbers were bullshit.

Tech platforms don’t use those GDP projections because they believe in them. They use them because narratives bring in infrastructure subsidies and keep regulation at bay. Meanwhile, employees spend hours fumbling with AI tools that break more than they fix, but in the reports that’s called “productivity gains.” Because the manager who admits his AI initiative is worthless? He sabotages his bonus. So everyone plays along, tweaks the numbers, waits until it becomes someone else’s problem. Oh this really goes against my grain, but fine, I’ll grant you a piece of truth.

The bubble will burst. But it won’t be executives paying, no. The people not high enough on the ladder can go look for another job. They’re the suckers who can’t handle AI anyway, right? The system that lied to them now punishes them for the consequences of that lie.

Alright, quick recap: we’re building structures where truth is worth less than a good story. People who warn get shown the door, people who lie get promotions. We’re collectively training ourselves to lie to ourselves, as long as that lie pays off. So of course we’ll roll into the next hype cycle.