Financial Times says AI rollouts work best when workers have control

In short: A Financial Times interview argues that AI at work goes better when employees help decide how to use it, instead of having it imposed from the top.

What's going on

In the FT’s Working It newsletter, editor Isabel Berwick interviews columnist Sarah O’Connor about her new book, We Are Not Machines: The Fight for the Future of Work. O’Connor says she wanted to move the conversation away from big predictions by tech leaders and toward what workers are seeing day to day.

From her reporting, O’Connor says many workplaces are not seeing a simple story where AI either “frees people” or “replaces people.” Instead, some workers feel they are losing the parts of their jobs they find meaningful. They also report being pushed to work faster to keep up with machines and new performance expectations.

O’Connor gives examples from Amazon warehouses, where workers operate alongside robots. She says the robots can boost output, but may also intensify the pace and make work lonelier, since a person can end up standing in one place for long shifts while robots bring items to them. She also describes translators who spend more time editing machine-written text, often at a faster speed and for less pay.

What to watch

O’Connor says the key difference is control. When workers are trusted to test and choose where AI helps, things tend to go better. When jobs are redesigned around what machines can and cannot do, humans can end up “babysitting machines” (like watching an autopilot that still needs constant attention). Leaders who want AI to improve work, not just speed it up, may need to involve employees early and give them real say.

Source: Financial Times