The Hidden Tax on Curiosity

Most people have more ideas than they pursue.

Not because they lack ambition. Not because they are lazy. Usually it is because the cost of exploring an idea has been high enough that they learned to filter aggressively before they ever began.

That is true for founders, operators, writers, marketers, product people, and probably anyone who has ever thought, Someone should build that, or There has to be a better way, and then gone back to work.

For a long time, I thought the biggest constraint on building was time.

It wasn’t time. It was friction.

More specifically, it was the hidden tax on curiosity.

The tax showed up early. Long before an idea had earned the right to become real, it already demanded a lot: time, coordination, context, confidence, and often other people. You had to push through ambiguity, survive the awkwardness of a bad first draft, and commit energy before you even knew whether the idea was good.

That changes behavior.

When curiosity gets too expensive, people do not stop having ideas. They just become more conservative about which ones they allow themselves to touch. They get practical. Rational. Selective. They ask whether something is worth the setup, the interruption, the meetings, the switching costs, the embarrassment of being wrong, or the effort required to drag a half-formed thought into the world.

Most ideas die there.

Not because they are bad ideas.

Because the cost of finding out is too high.

That is what feels different now.

What AI changed for me was not just speed. It changed the cost of exploration. It made it dramatically cheaper to turn a vague instinct into something tangible enough to react to. And once it becomes cheaper to find out, something important happens: more people start saying yes to things that would have previously stayed stuck in the category of interesting, but not worth the lift.

I am seeing that in myself, but I am also seeing it in the people I work with. They are exploring more ideas. They are taking more swings. They are trying things that would have felt too speculative, too cumbersome, or too marginal to justify six months ago.

That is not because everyone suddenly became reckless.

It is because being safe is getting more expensive than being daring.

The cost of being wrong has gone way down. You can sketch faster, test faster, model faster, discard faster, and refine faster. You do not have to overcommit just to learn. You can get to useful evidence earlier, which means more ideas deserve a shot than they did before.

That changes how people work.

But more than that, it changes how they feel.

Building makes me feel alive. It keeps me curious. It keeps me engaged with the world instead of just managing around the edges of it. For a while, too many ideas felt like obligations before they had a chance to feel like play. Every interesting direction came with enough friction that you had to decide too early whether it was “worth it.”

That is a deadening way to work.

You become so disciplined about protecting your time that you also start protecting yourself from the things that energize you.

That is the hidden tax.

And once I had language for it, I started seeing it everywhere.

Right now, I am working on a custom physical pet product using tscircuit. Six months ago, I would not have seriously considered something like this. Not because I was not interested, but because building a custom circuit board felt too far outside my natural range. I am not an electrical engineer. That used to matter a lot more than it does now. With AI, I can get much farther into the design process than I ever could have before. I may prove the concept and launch it this year, or I may not. That is almost secondary. The point is that the category became approachable.

The same thing is happening at BuySellAds. We are working on something in the programmatic ad space that would have been difficult to justify before. Not because the idea was bad, but because the economics were. The cost of training custom models at the scale we need to make an impact used to make the opportunity much less interesting. Now that cost is close enough to free that the equation changes. Things that previously looked too narrow or too operationally heavy to matter can suddenly become worth building.

I am also working on a website automation tool for myself to help manage all the different projects I enjoy juggling. Again, the problem was never whether the idea was useful. The problem was that it needs to interact with far too many APIs to have been a reasonable side project before. The integration burden alone would have been enough to kill it. Now it is not.

Those are three very different examples, but they all point to the same thing:

The world of things worth trying just expanded.

That does not mean you can say yes to everything. In fact, the opposite problem is starting to emerge. The number of things now worth exploring is close to limitless, which means judgment matters even more.

Because the easy part is moving earlier in the process.

The hard part is still the hard part.

Getting something all the way live. Getting it in front of real users or customers. Figuring out what people actually want. Earning traction. Earning adoption. Staying with something long enough to make it real.

And because more people can build now, more things will get built. Which means whatever you make will face more competition, not less.

That is why this is not a story about AI making execution irrelevant. It is almost the opposite. AI lowers the hidden tax on curiosity, but it raises the bar on execution.

Taste matters. Distribution matters. Persistence matters. Follow-through matters. The edge is becoming less about who can begin and more about who can carry something across the finish line.

But even with that caveat, the change is enormous.

For people who already liked to build, AI changes the shape of their ambition because more ideas survive long enough to be tested. But I think something even bigger is happening for people who previously assumed building was too far away from their natural abilities.

It does not feel so far away anymore, because it is not.

In a very real sense, the language of software is becoming natural language. English, Spanish, Dutch, Japanese — whatever language you think in can increasingly be used to turn ideas into working systems.

Step back and that is a wild thing to realize.

One of the biggest barriers between I have an idea and I can make something real is collapsing in front of us, and it is happening very fast.

That does not mean expertise disappears. It does not. The best builders will still be the best builders. The best engineers will still be the best engineers. But the surface area of who gets to participate has expanded dramatically, and it is still expanding.

A lot of people are going to describe this shift as a productivity story, and that is true as far as it goes.

But productivity is too small a word for what I think is actually happening.

The deeper change is that AI is reducing the hidden tax on curiosity.

And when curiosity gets cheaper, people change.

They become more experimental. More willing to try. More willing to risk being wrong. More willing to touch categories that used to feel outside their lane. Not because they suddenly became fearless, but because the old penalty structure no longer applies in the same way.

That is what I feel in myself right now.

Not certainty. Not mastery.

Energy.

I want to build more things. I want to test more instincts. I want to make more room for the part of me that comes alive when something moves from abstract to real.

Some of that work will fail. Some of it will never materialize. Some of it may turn into something meaningful.

That is fine.

Success is not the point, at least not at first.

Endless curiosity is the point.

For a long time, the hidden tax on curiosity was high enough that many people let go of more ideas than they should have.

I know I did.

I do not want to work that way anymore.

And I do not think I am alone.