The spiral trip of AI Integration

The AI layoff wave has turned. The rehiring has started.

The WSJ counted them yesterday: firms from CSX to Alphabet telling investors they'll add people, executives citing the costs and limits of AI, junior hiring switched back on. Robert Half puts numbers on it: 32% of hiring managers who cut for AI are rehiring those roles, and 55% admit the original call was wrong.

The market reads this as a comfortable story: AI was overhyped, humans won. Read the money instead.

• Step one: people were cut to fund AI, under real economic pressure. Margins compressing, tempo rising; the driver was never fashion.

• Step two: AI dropped into unredesigned processes returned nothing measurable; MIT told us that about 95% of pilots.

• Step three, now: rehire at today's salaries while the AI subscriptions keep billing.

And note what that adds up to.

Not the old position, something worse: the same efficiency as before the whole exercise, carried by a higher cost base, with part of the spend written off and the balance sheet thinner. The pressure that forced the first bet hasn't eased; it has grown, because the money for a second attempt is now scarcer and the price of getting it wrong again is higher.

This wasn't a cheap lesson. It was a loop that returns you below your starting point, wound tighter for the next turn.

Asked whether firms need more people or fewer, MIT's Paul Osterman gave the WSJ the honest answer: "We have no idea. No one has any idea." He's right; that is not a headcount question.

The problem is architectural: the operating model that makes people and AI produce what neither can alone.

Firms rehiring to go back will try to repeat the cycle, but that is a downward spiral.

Firms treating this as the second attempt at the redesign they skipped the first time have a different decade ahead.

That redesign is what Architecture of Intellect is about.

So which is your firm's rehiring: the way back, or the second attempt?