In a recent Guardian column, Gaby Hinsliff described how the promise that digital tools would free up our time has not been realised for many people, and warns that artificial intelligence may be amplifying that problem rather than solving it.
Hinsliff opens with a personal anecdote about an inbox swollen to more than 7,000 messages despite attempts at pruning, using it to illustrate a wider point: technologies introduced to save time — email foremost among them — have in practice created new kinds of busywork. Where a phone call or a quick conversation might once have sufficed, instant communications and bulk messaging have enabled organisations and individuals to send everything to everyone, generating a constant tide of material that must be sifted and deleted.
She argues that the same dynamic is emerging with AI. Governments and other institutions have publicly claimed that AI can transform the way the state operates, but Hinsliff says the reality is that large volumes of machine‑produced text and automated outputs are creating their own overheads. The result, she writes, is not only personal irritation but a measurable drain on time and resources as people and organisations attempt to filter, verify and manage a rising stream of low‑value content.
The column frames this as a policy as well as a practical issue: the benefits touted by proponents of rapid AI deployment need to be weighed against the operational costs of dealing with automated noise. Hinsliff suggests that the efficiency gains promised by new systems can be eroded by the effort required to police and curate their outputs, a burden that can fall on employees, public bodies and the wider public.
Her critique is part of a broader conversation about how to govern and implement AI in public services and private organisations. As adoption accelerates, questions remain about how to ensure that automation reduces workload rather than simply shifting or multiplying it, and about who should shoulder the costs of the extra labour required to manage machine‑generated material.