@NatureMC @susankayequinn I'm not sure a non-technical audience needs to understand the intricacies of methods used in science as long as they understand the scientific method and as long as it's followed.
Surely they don't understand how particle accelerators give us new insight into models of QM.
If you do need to explain things, use proper technical terms like statistical model, gradient descent, etc. Not bad anthropomorphizing analogies.
"AI" was already a *marketing term* in the 1950s. See: MIT's "AI lab".