I’ve noticed something strange lately: since AI became widely accessible, a lot of people around me have become noticeably more productive. But almost none of them feel less busy.

This shouldn’t be happening. If you can do the same things in less time, you should have more time left over. But that’s not how it’s playing out.

I’ve observed two very different situations, and both lead to the same place — more busyness.

For employees: fear is the engine

For a lot of people working in companies, the productivity gains from AI haven’t been particularly welcome news.

The problem is structural. Most employed workers sell their time to a company. When everyone’s productivity rises in sync, your relative position doesn’t change, and your salary doesn’t go up — unless you outrun everyone else.

This creates a strange dynamic: most of the gains from productivity tools flow to the company, not the employee. Workloads may actually grow (because more things become “feasible”), pressure increases, and compensation doesn’t keep pace.

So for most employees, the real motivation to use AI isn’t excitement — it’s fear. Fear of falling behind. Fear of being replaced. Fear that everyone else is using it and you’re not.

This is an underappreciated problem in the AI transition: when employees don’t directly benefit from the productivity gains they generate, they’re working inside a trap where standing still means losing ground. Productivity improvement becomes an arms race rather than a liberation.

How companies align employee output with company value — whether through profit sharing, clearer advancement paths, or redesigned roles — is something leadership needs to think seriously about. Otherwise, AI just amplifies the pressure rather than relieving it.

For independent workers: opportunity cost is the engine

I work independently, and my experience is entirely different — but the conclusion is the same. More busy.

There’s a clean economic logic behind it. Let’s say before AI, I could produce 100 units of value per hour. With AI, that becomes 300. But the opportunity cost of spending an hour resting or doing something else hasn’t changed.

Relatively speaking, working became more valuable — which means the cost of not working quietly went up too. Every hour of rest now forgoes 300 units instead of 100. The result? Work starts colonising time that used to belong to rest.

This explains something I’ve been seeing in the news: a lot of founders say they’ve stopped dating, or have cut down on almost everything that isn’t work. On the surface it sounds absurd. But the logic is entirely rational — every additional hour of work now generates far more value than it used to.

Of course, this logic has a prerequisite: your goals are open-ended. There’s no ceiling.

If you’re someone who just wants “enough” — a comfortable life, sustainable income — AI can actually make you more free. The same income in less time, leaving room for everything else. But for people chasing influence, impact, or unbounded achievement, the “income effect” never kicks in. The result is that you’re always working, and it always feels like it’s not enough.

Two situations, one outcome

What’s striking is this: whether you’re an employee or working for yourself, the productivity gains from AI seem to end in the same place — more busyness. The driving force is just different. One is fear, the other is opportunity cost.

This leads me to a more fundamental question: who are we being busy for?

For employees, much of the busyness is structurally imposed — there isn’t much room to opt out. But for independent workers, the busyness is often chosen, or more precisely, arrived at without consciously realising a choice was being made. The logic just carries you there.

The problem isn’t AI. It’s how we set our goals, and how companies design their incentives.

Productivity tools have never automatically produced freedom. They hand power to whoever decides how to use them — and that person still needs to know what they actually want.