A Camera in Your Living Room Is Cheaper Than a Manager

There’s a story doing the rounds about a woman in New Jersey who got fired after her employer’s “AI productivity software” photographed her, in her own living room, every three minutes. Not screenshots of her screen. Photos of her. Her home. Her couch, her bookshelf, whatever she happened to be wearing on a Tuesday afternoon. The quote that stuck with me, from someone commenting on it, was blunt and correct: a camera in your living room is cheaper than a manager.

I work hybrid. Some days in the office, most days at home, laptop open on the same desk where my daughter does her homework. So this one landed close to home, literally. I’ve spent two decades in IT and DevOps, and I’ve watched “how do we know people are actually working” go from a trust problem to an engineering problem. Somewhere along the way nobody stopped to ask whether it should be an engineering problem at all.

One comment on the thread brought up the Panopticon, Bentham’s old prison design where a single guard tower can see into every cell, but the prisoners never know when they’re actually being watched. The point of it wasn’t constant surveillance. It was the possibility of surveillance, all the time, which turns out to be just as effective at controlling behaviour and considerably cheaper to run. A hundred years later we’ve built the same thing into a $12-a-month SaaS subscription and called it a productivity platform. Bentham couldn’t get his version funded. Software has no such trouble.

What gets me isn’t even the technology, it’s the laziness dressed up as rigour. If a manager can’t tell whether someone’s doing good work without photographing their living room every three minutes, that’s not a monitoring gap, that’s a management failure. Deliverables, milestones, actual conversations about actual output: all of that still works, and it worked before anyone had a webcam pointed at their fridge. Replacing that with constant covert photography isn’t innovation. It’s outsourcing judgment to a camera because judgment is harder than installing an agent.

A few years back I turned down a contract that wanted keystroke logging and a mandatory always-on video call for the full shift. I didn’t need a Panopticon lecture to know that was a bad deal, it just felt wrong in the gut, the way a dodgy landlord inspection feels wrong. I’ve thought about that decision a lot since, mostly with relief, partly with the uncomfortable knowledge that not everyone gets to say no. The commenter who said they’d take the surveillance job anyway, because one paycheck beats zero, wasn’t wrong either. That’s the actual trap here. It’s easy to say “just don’t work somewhere like that” when you’ve got options. Plenty of people don’t.

We don’t have great answers for this in Australia yet either. The Fair Work changes around the “right to disconnect” were a start, but nothing in our law stops an employer from demanding photographic proof you’re at your desk, so long as you technically consented when you signed the onboarding paperwork nobody reads. Consent under those conditions isn’t really consent, it’s a toll you pay to keep eating.

I don’t think the answer is dragging everyone back to the office either, much as some executives would love to frame it that way. The honest answer is that trust is cheaper and more effective than surveillance, but it requires managers who actually know how to manage, and some organisations would rather buy software than train people. That’s a choice, not an inevitability. Worth remembering next time someone tells you the robots made them do it.