They are pretty smart, but the academics tend to live in their own world where they assume everyone else is just like them. I was working with one and sure, what they were saying was great academically and in theory. The problem became more apparent when you need to transform the theoretical into practice. Academics tend to add boxes and buttons randomly on a screen, and they expect the user to just know what the button does, or what it's for. And when you tell them their idea is retarded, they look at you like some peon piece of shit for not going along with their delusional ideals. Heck, sometimes the idea isn't even physically or practically possible, it works as an extension of another system, but on its own, it simply cant work without heavy review.
A great example, I was working on a piece of code. The academic wanted me to just add a button to add bounding boxes to a picture. I told him that doing so adds UI complexity which needs to be accounted for. He looked at me in disgust, as if such a basic task was beneath him. Of course, when it came time to implement, sure enough, we ran into problems getting the UI right. The reason was simple ... because what he was asking for was a collection of multiple screens, ui changes, and pretty much would require rebuilding the entire component.
At least they can sleep soundly knowing its someone else's problem to fix, not theirs
