When you keep presenting a supposedly unsolvable problem to people who don't know it, then you might get lucky. George Dantzig was one such person. He arrived late to class, saw some problems on the board, assumed they were homework, and solved them. He wasn't constrained by the common thought that these were "unsolvable".
They were not "unsolvable", they were simply not solved yet (at the time Dantzig saw them on the blackboard).
Did you read the comment? He said they were considered unsolvable at the time, not that they were unsolvable. Of course they're solvable - they were solved.
That is precisely my point. They were not considered "unsolvable" (i.e. impossible to solve), they were considered "unsolved" (i.e. not solved yet) and were presented as such. Do you really need me to explain distinctions in meaning here?
So I gave up and continued reading. Then somewhere in the 4th or 5th Chapter he says something like: Oh I hope you had fun with the Firing Squad Problem, I still work on it from time to time and hope to come up with a solution myself one day.
Facepalm.
Edit: Here is a link to the book, its enjoyable for experts and laymen alike. http://www.amazon.com/Feynman-Lectures-On-Computation-Richar...