It took ten minutes to build. Now we own it forever.

I want people to experiment.

Try the idea. Build the rough version. Find out whether it works. We can learn more from something we can touch than from another meeting about what might be possible.

But I’ve been thinking about a distinction that gets more important as AI makes building easier:

The freedom to experiment is different from the authority to give the whole company a new obligation.

Those can look identical on day one.

Someone builds a useful little tool. Adds a field. Creates a report. Introduces another step that solves a real problem.

It took almost no time. Everyone is impressed. YAYYYY CODEX.

Then other people start using it.

Someone has to explain what it means. Someone has to notice when it stops working. New employees have to learn it. Future changes have to account for it.

The person who built it moves on to the next idea.

“It only took ten minutes” describes the cheapest part!!

What makes this tricky is that the original idea can be good.

A report can be useful and still create more work than it removes. An extra option can help one person while making everyone else’s job harder. An automation can save time for its creator while giving somebody else a new thing to supervise.

You can accumulate a lot of individually reasonable additions and end up with a business that nobody can explain.

Now that people depend on something, removing it becomes its own project!

Now you need a conversation. A replacement. A transition. Somebody has to figure out what breaks.

The experiment has become a commitment, even though nobody remembers deciding to make one.

I don’t want to solve this by making people ask permission to have ideas.

That would destroy much of what makes these tools valuable.

I want more room to try things while they’re still easy to change. More willingness to throw away a first version. Less pressure to turn every promising experiment into the official way we work.

Sometimes you’re still figuring out the problem. Making the solution permanent too early just makes the next round of learning more expensive.

There should be a deliberate moment when we decide something deserves to stay.

At that point, the conversation gets bigger than whether it works. We need to understand who will own it, what it asks of everyone else, and whether the benefit is worth carrying that responsibility.

The builder can demonstrate the possibility but the organization still has to choose the commitment.

I also think leaders have to apply this to themselves.

A casual “could we add this?” carries weight when you’re the person in charge.

You think you’re brainstorming. Somebody else hears an assignment. A week later, there’s another thing to maintain because you had a thought on a Tuesday. (im guilty of this)

If I want the freedom to think out loud, I owe people clarity about when I’m exploring and when I’m making a decision.

Otherwise, my curiosity becomes their workload.

AI gives us the chance to test more ideas than we could before. I want to take full advantage of that.

But a successful experiment can end with learning something and deleting what you built. It doesn’t need to earn a permanent place in the business to have been worthwhile.

“Let’s try it” should leave us free to decide, later, whether we want to keep it.

Yallah Habibi,

Jon