Some people are really intelligent. They can look at a situation, come up with original creative solutions to problems, and cope with new and unexpected challenges.
Some people are really experienced. They've seen a lot of similar situations before, and they know how things should go by pattern matching the current situation to one in their past.
On the other hand, the intelligent people will often make beginner's mistakes, because they haven't seen it before, and nobody's smart enough to think of everything. The experienced people will often just lumber along, doing the same old same old, and not react to new information, or think outside the box to really change the rules.
I used to think intelligence beat experience, but I now believe neither one is actually superior to the other, and to make truly good decisions you must have both.
When I was at Microsoft, I felt like we often valued intelligence over experience—to a fault. This is where the "Microsoft takes 3 versions to get it right" thing comes from. We were all really smart, so we figured we could do anything, and people who had done similar stuff before were dinosaurs and we'd show them. The OpenGL versus Direct3D thing was an example of this, and I think it cost the game industry years of productivity. The SGI guys didn't come across as hyper smart in the argumentative Microsoft meeting mileu, so of course they weren't worth listening to. The industry paid the price.
Plenty of companies make the opposite mistake, where seniority and tenure win out over the better idea. Equally bad, obviously.
It takes a certain level of humility to value someone else's experience, especially if it's in a slightly different domain, but to do anything else is throwing away potentially valuable information that you're getting on the cheap. That would be pretty dumb, no?
It also takes a confident person to listen to some new cockamamie idea from a young smartass kid, even though you've done this same thing 20 times before and know the ins and outs. But who wants to do the same thing over and over? We sometimes need to try non-intuitive approaches to get out of local minima.