AI Doesn't Feel Like Real Work. Here's Why That's Actually Fine.
Published: March 29, 2026 | Category: 💡 Tips & How-Tos
There's a feeling a lot of AI users don't talk about out loud.
You've been using ChatGPT or Claude all week. You've generated more content in three days than you used to in three weeks. Your inbox is cleaner. Your proposals look sharper. By every measurable standard, you're more productive than you've ever been.
And yet something feels off.
You end the day with a weird kind of hollowness — like you did a lot but didn't do anything. You look at your screen and see five half-finished docs, three variations of the same email, a plan that's 90% drafted but somehow never sent.
A Reddit user named Scott_Weinberger put it perfectly this week: "It feels more like I'm managing drafts than doing work."
This is one of the most honest things anyone has said about working with AI. And it's happening to a lot of people.
What's Actually Going On
Before AI, finishing was hard. Writing an email from scratch took effort, so you pushed through it, sent it, and moved on. The friction was annoying but it was also a forcing function. You completed things because starting over was too painful.
AI removes that friction entirely.
Now you can generate a draft in 10 seconds. Not happy with it? Regenerate. Tweak the prompt. Get five more versions. Compare them. Mix and match. Go back and try a different angle.
This isn't laziness. It's actually a reasonable response to having infinite drafts available. The problem is that infinite drafts with no finish line creates a new kind of work that feels productive (you're busy, you're iterating, you're improving things) but doesn't ship anything.
Productivity researchers call this task expansion — when a task fills all available time and cognitive energy because there's no forcing function to end it. AI has given us a powerful new form of task expansion.
The solution isn't to use AI less. It's to understand the new mental model.
The Shift: From Effort to Outputs
Before AI, your mental model of work was effort-based. You knew a proposal "counted" because you spent four hours on it. The time proved the value.
AI breaks that model completely — and it needs to be replaced, not patched.
The new model is output-based: did something ship, or didn't it? That's the only metric that matters now. How long it took is irrelevant.
This sounds simple, but it requires rewiring a lot of deep assumptions about what work is. The guilt that AI users feel — about finishing too fast, about delivering "too easily" — comes from still running the old operating system while using new tools.
You're not cheating. You're just faster. The value you deliver is the same or better. The time you saved is yours to keep.
How to Actually Finish Things with AI
If you're stuck in draft-management mode, here are three rules that help:
1. Set a draft limit before you start. Before you open ChatGPT, decide: "I get two drafts of this. I'll pick the better one and send it." The moment you decide a limit up front, the infinite-draft trap closes. One or two iterations is almost always enough — you're just not giving yourself permission to stop.
2. Build a "good enough" trigger. AI outputs don't need to be perfect — they need to be sent. A useful mental trigger: "Would I be embarrassed to send this?" If no, it's done. If you find yourself on your fourth version trying to make something slightly better, you're past the point of diminishing returns.
3. Separate generation from editing. The loop that creates fake progress is: generate → tweak → regenerate → tweak. Break it by separating the sessions. Use AI to generate. Then close the chat, open a document, and edit manually. Once you're in editing mode, you can only improve what's there — not start over.
The Feeling Doesn't Fully Go Away
Here's the honest part: even when you nail the workflow, there's still a low-grade strangeness to finishing a solid piece of work in eight minutes. It doesn't feel like you "earned" it yet.
That feeling is normal. It's a cultural hangover from decades of equating effort with value.
But ask yourself: does your client care how long it took you? Does your customer care how many drafts you went through? Does the proposal that wins the business feel less valuable because you wrote it in 20 minutes?
The work is the work. The result is the result.
AI changed how the sausage gets made. It didn't change what good sausage tastes like.
Give yourself time to update your operating system. Most people are about six months behind on that adjustment.
Have a question about working smarter with AI? Reply to this article — we read everything.