What if the problem isn't too few jobs?

The consensus is tidy and it is everywhere, including in this very edition: AI is hollowing out entry-level work, hiring is down 14%, and the state must rush to catch the displaced. Taken at its strongest, it's a serious case — the British Chambers of Commerce found UK AI adoption at 54%, and among the firms going deepest, roughly one in five report cutting staff. Something real is happening. I'm not going to pretend otherwise.

But here is the case nobody made this week. Britain's defining economic disease for two decades has not been a shortage of jobs — it has been a shortage of productivity. We work long hours to produce less per hour than the Americans, the Germans, the French. If AI does what its boosters claim and lets one person do the work of three, that is not, for a chronically unproductive economy, a catastrophe. It is the first plausible cure anyone has offered. The Bank's own governor has said AI will both create and destroy jobs; the OECD thinks it could add over a point to annual productivity growth.

Concede the obvious: that transition will hurt, and the school-leaver in 2028 doesn't care about aggregate productivity if their rung is gone. The pain is real and Elena is right to track it. But notice what the week's framing quietly assumes — that the goal is to preserve the old shape of work. What if the prize is to let it change, and to fight like hell over how the gains are shared rather than over whether the machine is allowed to help? The jobs panic is a 20th-century worry. The 21st-century question — who owns the productivity AI unlocks — is the one Westminster, and most of us, are too nervous to ask.

I might be wrong. The productivity dividend may never arrive, and we'll have traded good jobs for nothing. But if I'm right, we are spending this week frightened of the wrong thing — guarding the lifeboats on a ship that was finally about to find its engines.