This is quite interesting:
It's the James Burke institute which is related to James Burke who wrote the famous 'Connections' documentary from back in the 70s.
Basically, they're trying to build webs of connections between things, and plotting them in 3D, so you could pick something, like the computer and trace backwards and find out what things led to its creation.
They're doing it the hard way, but it struck me that the wikipedia might be mined for this kind of thing- that many connections may already be there and that the dates contained in articles might allow a creation of an interactive graphic for looking at the wikipedia in a new way.
It also struck me that perhaps the wikipedia doesn't value antecedents very highly. I think that history sections tend to cover the first example of something, but not so much things that lead up to it, that weren't it, or forces that helped create it or make it practical or economic to do that way.