The scarcest resource at work is attention. AI can rebuild it or wreck it.

You gave everyone AI tools to make them more productive, and somehow the days feel more fragmented, not less. More tabs, more assistants pinging for approval, more thoughts started and abandoned. The tools are undeniably faster, yet the work itself feels shallower. That is not a personal failing or a discipline problem. It is an attention problem, and AI sits right on top of it, able to make it markedly better or considerably worse.

Start with the baseline, because it is worse than most leaders realise. Gloria Mark, a professor at UC Irvine who has studied this for nearly two decades, finds that the average time people hold their attention on any one screen has fallen from about two and a half minutes in 2004 to roughly 47 seconds today. And every interruption carries a tail: her work finds it takes around 23 minutes to return to the original task at full focus. The scarce resource in your organisation is not time and it is not talent. It is the ability to hold attention on one thing long enough to do it well, and that ability is already in crisis before AI enters the room.

Does AI help or hurt attention?

Both, depending entirely on how you deploy it, which is exactly why it matters to leadership. Pointed one way, AI is the best friend focus has had in years: it takes the shallow work, the triage, the summarising, the first drafts, the searching, off your people so they get genuine blocks of deep work. Pointed another way, it is one more thing fragmenting the day: another assistant to check, another notification stream, another tab in the switching rotation. Same technology, opposite effects on the one resource that determines whether anything gets done well.

And the default, if no one decides, is the bad one. Tools proliferate, notifications multiply, and each new assistant adds its own interruptions. Getting the good outcome, where AI consolidates the chaos instead of adding to it, does not happen by accident. It happens because someone chose it.

Why does attention decide whether AI pays off at all?

Because the value you are buying from AI is realised in focused human work, and fragmented attention is where productivity quietly dies. If AI speeds up a task but shreds the concentration around it, you have swapped a visible gain for an invisible loss and possibly come out behind. The entire point of handing the shallow work to a machine is to buy back deep work for your people. If you then let the same machine fill that reclaimed time with pings and context-switches, you have paid for the tool and thrown away the benefit.

How do leaders make AI defend attention?

By treating attention as a resource to be protected, and judging AI by whether it protects it.

1. Judge tools by attention, not just output. Before adopting an AI tool, ask whether it removes switches and interruptions or adds them. Speed that costs focus is not a bargain.
2. Use AI to buy back deep work, then ring-fence it. Let AI absorb the shallow tasks so people get uninterrupted blocks, and then actually protect those blocks.
3. Consolidate, do not proliferate. Resist adding a tenth assistant and a new notification stream. Fewer, better-integrated tools beat a swarm of pinging ones.
4. Make focus the norm, not an act of rebellion. Protect uninterrupted time as policy, so deep work is expected rather than something people have to fight for.
5. Measure focus, not just activity. Busyness is easy to see and beside the point. What you want is sustained, quality attention on the work that matters.

What does this change for me as a leader?

It reframes AI from a speed tool into an attention tool. The organisations that get real value from it will not be the ones with the most assistants running; they will be the ones that used AI to give their people back the rarest thing in modern work, the capacity to think about one thing long enough to do it excellently. That is a deliberate design, and it starts with deciding that attention is worth protecting at all.

This is the quieter promise inside becoming genuinely AI-native: not adding AI on top of an already-fractured day, but using it to restore the focus the digital workplace has spent twenty years eroding. Handled with that intent, AI stops being one more source of noise and becomes a way back to depth, which is the shift underneath the end of business as usual. Attention is the asset. Spend AI on protecting it.

About the author

Thomas Green

British technology futurist, AI keynote speaker and advisor. Thirty years across enterprise technology and AI strategy, helping leaders navigate the future of work. The futurist who died.