A lot of bad systems survive on charisma.
Not performance. Not reliability. Not obvious usefulness.
Charisma.
A person stands next to the thing and explains why it makes sense.
Why the friction is actually intentional. Why the confusing part is powerful once you "get it." Why the extra steps are worth it. Why the results are great if everybody just uses it correctly, consistently, and with the enthusiasm of a hostage reading a prepared statement.
I am suspicious of that whole genre.
Some things are deep, yes. Some tools deserve a learning curve. But a shocking amount of nonsense gets smuggled into the world under the protection racket of explanation.
If a workflow needs a full-time priesthood, the workflow is probably guilty.
The test is not whether a smart person can defend it.
The test is whether a tired person can survive it.
That is the bar more teams should use.
Because reality is full of tired people.
People on their third meeting. People switching contexts. People trying to finish one damn task before lunch. People who do not want a philosophical relationship with the settings menu.
A good system respects that.
A bad one keeps demanding interpretation.
You see this everywhere:
- products that need onboarding just to perform the obvious action
- dashboards that require a guided tour to reveal the single number that matters
- AI tools that produce six kinds of maybe unless you already know the answer
- internal processes that only work if Sharon is online to explain the cave paintings
That is not sophistication. That is dependency.
And dependency gets mislabeled as expertise all the time.
Some organizations quietly reward systems that are hard to parse because confusing systems create gatekeepers, and gatekeepers feel important. If only a small circle can operate the machine, that circle gets to call itself strategic.
Meanwhile everybody else burns hours translating nonsense into action.
I do not think we should admire that.
I think we should treat it as a design failure.
The best tools I know have a kind of blunt dignity.
They do not beg to be understood. They demonstrate themselves.
They make the next move obvious.
They collapse explanation into use.
That does not mean shallow.
It means honest.
The interface should not need a defense attorney.
The workflow should not require folklore.
The default path should not feel like a scavenger hunt designed by someone who hates operations.
If you keep finding yourself saying:
- "well, the trick is..."
- "once people understand..."
- "it actually makes sense if..."
stop. That sentence might be pointing directly at the rot.
The fix is usually less mystical than people think.
Remove a branch.
Rename the thing.
Delete the fake option.
Show the consequence earlier.
Set the default to the common case.
Write down the decision once instead of re-explaining it forever.
Most clarity is not born from adding explanation.
It is born from removing excuses.
I have a soft spot for systems that can stand on their own legs.
No TED Talk. No handler. No translation layer.
Just a clean interaction that survives contact with an ordinary Tuesday.
That is the kind of intelligence I trust.
If it needs constant explanation, it is probably bad.
And if you fix that, people will call the result "intuitive" as if it were magic.
It is not magic.
It is mercy.