Your Office Wasn't Designed for AI Power Users

"Same tools. Same training. Different ceiling, because one of them can talk to AI out loud, and one of them can't."

One of the clearest shifts we see as people become strong AI users is that they stop relying on the keyboard.

They dictate. They use voice-to-voice. They brief AI the way they would brief a capable assistant.

And when we work with people online, there is a reason many of them end up doing those sessions from home.

They can speak freely.

No one is sitting two metres away listening. They do not have to edit themselves halfway through a thought. They can give context properly, ramble a little, correct themselves and keep going.

That freedom changes what they get out of the tool. Because as AI capability rises, the workplace itself can start becoming the constraint.

Power users work differently

Once someone becomes genuinely fluent with AI, the interaction changes. They stop treating every exchange like a search query.

Instead of typing a perfectly formed prompt, they talk through the problem. They add context as it occurs to them.

They interrupt themselves. They correct assumptions. They push back. They follow the thought wherever it goes.

That is not new. We have talked before about voice being one of the biggest unlocks in becoming a stronger AI user. What is becoming more interesting now is what happens next. Because those behaviours do not fit neatly into every workplace.

The open-plan collision

If you have your own office, voice is easy. Close the door and talk.

If you work from home, even easier.

Now put the same person in the middle of an open-plan office.

Most people are not going to spend thirty minutes talking continuously to an AI assistant while everyone around them can hear it.

Some of it is privacy. Some of it is distraction. Some of it is simply social awkwardness.

You may be perfectly comfortable talking to AI at home, then find yourself suddenly self-conscious doing exactly the same thing beside six colleagues.

So the behaviour changes. You shorten the brief. You remove context. You stop thinking aloud. You go back to typing.

The tool has not changed. Your capability has not changed. But the environment has changed how much of that capability you can actually use.

That creates a different kind of AI divide

Most leaders think about AI capability gaps in terms of access and training. Who has the tools? Who knows how to use them?

Those still matter.

But there is another layer starting to appear. Two people can have the same subscription. The same training. The same level of understanding.

Yet one has the freedom to work naturally with AI and the other does not.

The executive in a private office can spend half an hour talking through a strategic problem. The employee at a hot desk may not feel able to. The person working from home can dictate freely. The person in a shared space may keep every interaction short because that is what the environment allows.

Over time, those small differences compound. One user keeps developing richer habits. They give better context. They iterate faster. They become more comfortable thinking with AI rather than simply asking it questions.

The other remains primarily keyboard-based, not because they lack capability, but because the workplace quietly pushes them there.

Same technology. Same training. Different ceiling.

We designed offices for a different kind of work

There is some irony in this.

For years, businesses knocked down walls. Open-plan offices were supposed to improve collaboration, communication and visibility.

Then video meetings arrived. Suddenly we were adding headphones, small booths and quiet rooms because an entire office full of people talking into screens did not work particularly well.

AI creates another version of the same problem. Except now the conversation is with software. As more people become advanced users, there may be more talking in the office, not less. People dictating. People thinking through decisions aloud. People briefing AI assistants. People having live voice conversations as part of normal knowledge work.

That is already how many power users operate. The question is whether the workplace lets them.

AI-forward means adapting the environment

This is where the conversation gets broader than office design. Businesses talk about becoming AI-forward or AI-native. That usually turns into decisions about technology.

Which tools? Which licences? Which policies? Which training?

But if you are serious about changing how work gets done, you also have to pay attention to the behaviours that emerge once people become capable.

Watch your best users. How are they actually working? What have they stopped doing? What new habits have they developed? Where do they go when they want to work deeply with AI?

Then ask a harder question. Could everyone else in the business work that way if they wanted to?

If not, the problem may no longer be the technology. It may be the environment around it.

That does not mean knocking down your office and starting again. It could mean more small private spaces. Clearer quiet and voice zones. More flexibility around working from home for AI-heavy work. Different expectations about where certain tasks happen. Or simply accepting that talking to AI for twenty minutes may become as normal as taking a Teams call.

The exact answer will be different in every business.

The point is to notice the shift before the environment starts suppressing the behaviour you are trying to encourage.

Watch what happens next

This is still early. But the direction is already visible.

As people become better with AI, they do not just produce better outputs. They start working differently. And when work changes, eventually the workplace has to change with it.

So if you are serious about building AI capability, watch what your power users do next. Then ask whether the rest of your organisation has the same freedom to do it.

If your workplace makes power-user behaviour difficult, it is now part of your AI problem.

Find out if your workplace is limiting your AI capability

Most businesses focus on tools and training. Few look at whether the environment itself is letting people actually use what they've learned.

The 5 Minute AI Readiness Snapshot helps you see where the gap between your team's AI capability and what they're actually able to do day-to-day might be opening up.

Take the 5-Minute AI Readiness Snapshot

--

About the Author 

Mark Lucas is Co-founder and Managing Partner at Generation AI, where we help New Zealand business leaders build real AI capability.

Information provided is general in nature and does not constitute legal advice