The office isn't dying to AI; it's evolving

The more executives insist the office is sacrosanct, the more the tools they worship may make that insistence pointless. Fast Company’s “How AI could kill the return to office” stakes out a clean headline: AI could keep people remote by changing how work gets done. I’ll be honest: that’s a real possibility. But treating AI as the decisive factor misses the messier part of the story—who controls the tools, and who benefits from where bodies actually sit.

Office real estate: mausoleum or museum?

Here’s the thing: generative AI can strip out a lot of the original justifications for dragging everybody into the same building. If AI agents can synthesize meeting notes, triage email, draft proposals, and simulate stakeholder reactions asynchronously, you don’t need ten people physically around a whiteboard at 10 a.m. to move a project forward. That’s not fantasy; it’s a clear technical trajectory that Fast Company is right to flag.

But office towers were never just about productivity.

They’re status symbols, proximity machines, and architectural org charts. A corner office is a power diagram in glass and steel. Even if AI makes knowledge work highly distributable, landlords and companies won’t simply surrender that. They’ll rebrand.

Think less “rows of desks” and more “curated collaboration hubs”: fewer seats, more staged serendipity, very expensive coffee. The value shifts from generic square footage to tightly managed access—executive reviews, sensitive briefings, high-stakes client meetings. The physical office turns into a museum of presence: exclusive, expensive, and deeply political. You’re not commuting to work; you’re commuting to a performance.

That’s where Fast Company is most convincing: AI absolutely lowers the technical need for co-presence. But technical feasibility doesn’t dissolve status incentives, and it definitely doesn’t erase sunk real-estate costs that sit on balance sheets like stubborn ghosts.

The hidden axis: control of the AI

Funny thing is, amid all the talk about remote work and digital tools, hardly anyone wants to squarely address who actually controls the AI.

You’ve got two forces pulling in opposite directions:

• AI as an autonomy engine that lets individuals work from anywhere.
• AI as a surveillance and gatekeeping engine that recentralizes power.

Employers can use AI to monitor productivity, parse communication patterns, and surface who gets what kind of work. That’s the polite version. A more pointed one: they can quietly reward those who show up in person with better training, more sensitive data access, or faster AI assistants that run against on-premises infrastructure. Same technology, very different power dynamics.

Fast Company hints that work will be restructured. It underplays how hierarchies can be re-wired.

Consider the location of the most capable AI stacks inside a large company. If the best-performing systems sit behind corporate firewalls and are wired into proprietary datasets, then proximity to those systems—and to the humans who grant access—starts to matter. The office becomes a choke point for privileged AI services, tugging some workers back even as others stay remote.

Call it the Neuromancer paradox: cyberspace makes distance irrelevant, until the systems you need are guarded by gatekeepers. William Gibson’s world runs on access as status; AI-powered workplaces can easily drift down the same path.

Two consequences follow from that logic.

First, inequality inside firms can widen. Those with in-office access to the most capable tools get better assignments and faster promotions. Remote workers risk getting the leftovers: thinner models, slower support, less visibility.

Second, cities fragment. Certain districts will cluster elite hybrid hubs—AI-intensive offices, satellite studios, private collaboration spaces—while other areas hollow out. Fast Company nods at big-picture urban shifts, but not at how the same AI stack can create internal “classes” of workers on different rungs of the access ladder.

The counter-argument—and where it breaks

The neat rebuttal is simple: executives will just mandate return-to-office and that’ll be that, AI or not.

Sure, but mandates don’t exist in a vacuum. When workers can credibly perform from anywhere, and when AI makes that performance measurable, heavy-handed enforcement has a cost. Attrition is a quiet but powerful negotiator. You can threaten careers; you can’t force people to ignore every competing offer.

You also run into the geometry of cities. If transit, childcare, and housing don’t line up, blanket mandates turn into daily friction. AI doesn’t fix infrastructure, but it does expose how unnecessary some commutes are by showing that output doesn’t fall when people stay home.

That doesn’t mean employers are powerless here. Far from it.

They can redesign roles so that the highest-impact, highest-visibility work “just happens” to require co-location. They control the evaluation systems that decide whose output counts most. They decide whether the AI stack boosts individual autonomy or tightens managerial oversight. Fast Company is right that technology enabling remote work is a necessary condition for lasting change; it’s just not a sufficient one. Power, incentives, and architecture still arbitrate who gets to opt out of the train ride.

A quick detour to history

If this script sounds familiar, it’s because we ran a version of it with email and early internet tools. Back then, we were promised “work from anywhere” too. What we actually got was BlackBerry culture—constant reachability, plus offices that didn’t disappear so much as mutate. The tools made distance less of a blocker, but control stayed remarkably centralized.

AI has more raw horsepower than email ever did, but the pattern rhyme is hard to ignore.

So here’s what I’m watching: where companies physically locate their best AI environments, how promotion criteria subtly shift for remote versus in-office staff, and how landlords pitch their next generation of spaces. Those decisions will quietly answer Fast Company’s headline: whether AI really kills the return to office, or just teaches the office a new trick.