Everyone I talk to is asking the same question about AI and work. Which jobs will it take? I think that is the wrong unit of measurement, and one of the world's leading economists on this subject keeps pointing at the right one without quite naming it.
Stanford's Erik Brynjolfsson has argued for years that the real story is task reorganization, not job replacement. Companies gain the most when they rethink how work is arranged around the technology, because a job was never one thing. It is a bundle of tasks, and AI takes some tasks while leaving others untouched. His lab went further and asked 1,500 workers across hundreds of tasks how much human involvement they actually want when AI agents arrive, and built a scale to measure it.
Here is what I see when I read that research through twenty years of enterprise work. A job is a bundle of tasks, and a task is a bundle of decisions. When people say an agent took over a task, what actually happened is smaller and more precise. The agent took over certain decisions inside that task. Which refunds to approve. Which orders to release. Which cases to escalate. The work that moved was deciding, and deciding is the one thing companies have always managed carefully, in writing, with names attached, when humans did it.
What actually transfers when work moves to an agent?
Not the whole job, and usually not even the whole task. What transfers is authority over specific decisions, and that transfer is happening today mostly without anyone signing it.
AEBOK holds that work moves to AI one decision at a time, not one job at a time, and that every transfer needs the same paperwork we have always given people: what may be decided, where it must stop, and which named person answers for the outcome.
Those workers in the Stanford study, saying how much involvement they want to keep? They are voting on decision rights. They just have not been given the vocabulary, because their companies have not written the page.
The question to bring to your own role
Do not ask whether AI will take your job. Ask which decisions inside your work are being transferred, who approved the transfer, and what your job becomes once the deciding is shared. People who can answer those questions are not waiting for the future of work. They are managing it.
Which decision in your own work do you think moves first? I would genuinely like to hear it.
Rabi Jay is the founder of The Autonomous Enterprise Institute and author of books on enterprise technology across the ERP, cloud, and AI eras. He writes The Autonomous Enterprise newsletter and hosts The Road to the Autonomous Enterprise, a daily show on AEBOK, the management standard for delegating decisions to AI.