2026-04-15 · #1

Why AI work can feel oddly exhausting

— It might not be workload, but decisions

You use AI, the work moves forward, and yet something feels off. You’re done — but your mind stays noisy.

If you’ve used tools like ChatGPT or Claude, you may have felt this:

  • You’re making progress, but you’re strangely drained
  • Your mind gets hazy and you can’t decide what to do next
  • The next day, the fatigue is heavier than you expected

“It’s supposed to be convenient — so why am I this tired?” It’s a reasonable question.

In 2026, Harvard Business Review introduced a phrase for a similar pattern: “AI brain fry.” The idea is simple: when you juggle multiple tools and keep monitoring, verifying, and correcting outputs, your attention can saturate. The discussion is based on a BCG study of 1,488 U.S. workers at large companies.

AI work isn’t always “easier” — it can add decisions

AI can speed things up. But the faster it gets, the more your job quietly shifts.

For example, you start stacking micro-decisions like these:

  • Which answer do I pick?
  • What should I rewrite?
  • Is this direction safe?
  • Should I ask again, differently?

Each one is small. Do it dozens of times in a short window, and it starts to bite.

So AI work can be not only “faster,” but also a chain of repeated decisions.

Ads
Ads may be shown to support the site.

The real driver: decision density

What matters isn’t just how much you did, but how tightly decisions were packed. Let’s call that “decision density.”

  • Comparing multiple AI tools
  • Iterating by trial and error
  • Switching between projects or tasks

More options create more decisions — and more decisions increase cognitive load.

Feeling hazy is a normal signal

When decisions pile up, your brain can get stuck holding a queue of “pending” items. That’s when these show up:

  • Thoughts won’t organize
  • Your judgment gets sloppy
  • You can’t find a stopping point

That hazy feeling isn’t laziness. It can be a sign your brain is hitting its processing limit. And if you push through, you may “finish” — but pay for it tomorrow.

The point isn’t that AI is bad. It’s that the way you use it can add load — especially constant checking and switching.

Quick moves (short)

  • Cap “one more try” by time, not by attempts (e.g., stop after 10 minutes).
  • Pick one answer to move forward, then switch tasks.
  • After switching tools, write one line: “what I’m doing right now.”

Try it now (rough guide)

Try it now (rough guide)
Let’s take a quick look at how much load you might be carrying right now. Use it as a cue to consider a break or stopping for today.
  • Use the result as a rough guide for decision fatigue.
  • This is not a medical or psychological diagnosis.