Learning With Leo: Where Reliability Actually Hides
Knocking Down the Bone Pile: X-ray Inspection of Ball Grid Array Solder Joints
Last month, I made the case that you should not buy AI in the hope that it will fix the business. You need to learn it yourself first.
If your AI is failing, it’s not because it is lazy or broken. It is failing because it does not know your plant. It does not know your quoting rules, how material gets released to the floor, what your customer expects in an 8D1, or which spec outranks another. It certainly does not know where your team hides the real answer once the formal procedure and reality have drifted apart.
AI is like a smart new engineer on day one, bright and fast and generally informed, but with no idea of your customers, process names, capability edges, handoffs, approval rules, or evidence chain. If you want useful output, you have to teach it those things.
The Recognition Moment
You have probably seen this before. You ask AI to draft a customer response and it sounds polished, but overpromises. You ask for a material shortage note, and it misses the one fact the floor needs. You ask for an 8D, and it hands back a tidy report that sounds finished before the evidence is in.
You gave it a task but never taught it your process. In a PCB shop, that matters more than people think. Our work is context-heavy, document-heavy, and exception-heavy. A generic answer can sound good while being wrong for your plant.
To continue reading this article, which appeared in the July 2026 I-Connect007 Magazine, click here.
