Fake Work Was Never About Laziness

I used to think fake work was a personality problem — some people are hustlers, some are slackers, and the slackers fake it. I don’t believe that anymore.

Fake work is structural. It’s not something a bad employee does. It’s something every knowledge job produces, by default, because the value of knowledge work is almost impossible to measure directly. So organizations measure something else instead — reports written, decks polished, meetings held, code shipped — and reward that. The proxy replaces the thing it was supposed to stand in for.

I recognize this pattern from my own career. Managing a team means spending most of your time translating: turning a boss’s strategy into a team’s tasks, turning a team’s output back into a report a boss can read. You get good at that. People like you for it. And you can go years without asking the only question that matters: did anything actually change because of what I just did?

That’s the real test. Not “did I produce something,” but “did the next decision look different because this existed?” If a report just proves the team is busy, it isn’t work — it’s an alibi.

The dangerous part is how this rewires you without asking permission. A day spent in meetings, replying to email, formatting slides feels productive. Your brain tells you: today was a good day. But feeling busy and creating value are two completely different signals, and if you’ve spent eight years being rewarded for the first one, you lose the ability to even notice the second one is missing.

Here’s what changes now: AI is extremely good at producing the proxy. A polished strategy memo used to take a team a week — that scarcity is exactly what made it valuable-looking. It signaled effort, and effort used to correlate, loosely, with value. AI breaks that correlation instantly. The memo still gets produced. The effort behind it drops to zero. And once the cost collapses, the price should collapse with it — except most organizations don’t yet know how to stop paying for it.

So the honest short-term forecast isn’t “fake work disappears.” It’s “fake work explodes first, then gets exposed.” More reports, more meetings, more shipped code — because AI just made all of that cheaper to produce, and cheap things get overproduced. Then, slowly, the mismatch becomes impossible to ignore: nobody asks anymore whether a decision changed, they just ask if a decision changed, and it doesn’t. That question was always available. It just used to be too expensive to keep asking, because everything else was expensive too.

I don’t think the right response is to panic or quit your job tomorrow. It’s to start noticing, in your own work, the moment you’re producing evidence instead of outcome. Ask the boring question every time: if this is done, what actually happens next that wouldn’t have happened otherwise? If the honest answer is “nothing, but now people know I did something” — that’s the tell.

What actually matters, stripped down to three things: define the problem — know what you’re really solving before you start solving it. Judge it — decide whether it’s actually worth doing, not just doable. Own the consequences — be accountable for what happens after, not just for having produced something. These three are not automatable, and they were never fake. They were just underpriced, buried beneath the loud, visible, proxy-generating work that was cheaper to reward and easier to see.

I’d call this less a threat and more an overdue correction. The fog of expensive-looking fake work is burning off, and what’s left is exactly these three things — define, judge, own. That’s not something to fear. It’s something to get ready for.