the following site illustrates how the delays between successive
radical breakthroughs in computer science decrease exponentially: each
new one comes roughly twice as fast as the previous one:
http://www.idsia.ch/~juergen/computerhistory.html
cheers,
Juergen