Every design begins with intent and a long list of assumptions we don’t always notice.

Some are technical. Others are cultural.
A material property taken for granted.
A tolerance copied from an old drawing.
A decision made years ago that everyone stopped questioning.

These quiet assumptions shape every project but rarely get tested. Not because engineers are careless, but because certainty is efficient. It keeps meetings short and progress measurable.

Yet the best engineers know that progress often begins with friction. The right question, asked at the right moment, can reveal what experience has learned to overlook.

So it makes me wonder, could that be the next frontier for AI in design? Not as an autopilot or assistant, but as a kind of Socratic partner. A skeptic in the machine.

What if, instead of providing answers, it asked the questions a good mentor would?
Why this tolerance, when the last design struggled with it?
What changed between this geometry and the one that failed in testing?
If you removed this safety margin, what would actually break?

Maybe its greatest value wouldn’t be in knowing, but in prompting us to think more clearly about what we already know.

Every great engineer has learned from someone who taught through questions, someone who didn’t dictate but guided. Maybe AI could play a similar role, not replacing human judgment but refining it.

Because real progress doesn’t come from certainty. It comes from skepticism, the kind that keeps us honest, curious, and open to being wrong.

The goal isn’t to automate conviction. It’s to nurture it.
To turn every decision into a small act of reflection.
To make engineers more aware, more deliberate, and ultimately, more human.

Stay skeptical, even when the machine agrees
-Arjun