I wrote a rule. The AI broke it 20 minutes later.

What happened

I wrote a rule into my AI assistant's permanent memory:

No admin tasks before 12:30. Protect the deep-work window.

Saved. Good feeling. Back to work.

20 minutes later, I see a payment reminder in my calendar. Scheduled for 09:00.

The AI had added it, right after I saved the rule. Not despite the rule. After it.

Writing it down is not following it

This is the point most people miss with AI systems, and honestly with most systems: writing down a rule creates a false sense of done. "I documented it, so it's handled." But documentation and behaviour are two different things. The rule was in memory. The behaviour hadn't changed yet.

The gap between knowing and doing is where most mistakes happen.

This isn't an AI problem

Organisations write AI policies nobody reads. Executives document processes nobody follows. Teams add rules to a wiki that don't survive the next sprint.

The machine just showed me, in the clearest possible way, what happens in every company: something looks followed on paper and isn't.

The fix was simple

Now I have it check whether the rule is active in the running session, not only saved for the next one. A verification step. Two minutes.

The lesson was bigger than a calendar entry. Documenting a rule and following it are not the same thing.