The workplace debate about artificial intelligence in 2024 was dominated by a single anxious question: how many jobs will be lost? The forecasts ranged from apocalyptic to dismissive, and the argument consumed boardrooms and editorial pages alike. But it may have been the wrong question from the start.
Watch how AI is actually being used in a modern office and a different pattern emerges. AI drafts, summarises and suggests; it does not, yet, decide. The disruption is happening not at the level of jobs but of trust, in who is allowed to write the first draft, who is credited with the thinking, and who is accountable when a machine-generated answer turns out to be confident and wrong.
The reorganisation that matters is subtler and more corrosive: an erosion of the assumption that human judgement, verification and accountability sit at the centre of professional life. The task for the next decade is not to decide whether machines take jobs, but to decide how much of our institutional life we are willing to hand to systems that cannot be responsible for what they say.