tl;dr: There was a prize offerered in the 1950s-1960s for the first person to build a human-powered plane that could run a short course (call it the X-prize of human flight). Lots of people tried to win by building planes that looked like planes.<p>Paul MacCready was the first to realize the key to the spec: it had to fly, but didn't have to look like a plane. He rapidly prototyped things that flew but didn't look like planes, and used substitutes for normal safety procedures (flying low and slow rather than building a strong cockpit). He learned from many failures rather than spending months building to learn once.<p>Lesson: make sure you understand what the spec is really asking for, and prototype quickly around that.
I remember reading about the follow-up, the Gossamer Albatross, when I was a kid and being enthralled. They flew it across the English Channel to claim the second Kremer prize.<p>Quite apart from the point about the advantage of inexperience, the Gossamer planes are wonderful creations.
I recently read "More With Less", Paul Ciottis biography of Paul MacCready; if you want to dive deeper into the fascinating story of the creation of the Gossamer planes (starting with MacCready's unique background and network, the team re-shuffles [and alienations], tons and tons of iterations, all the downs for several very big ups; sounds like a start-up, doesn't it?), I certainly recommend it.<p>It required a man of extraordinary skills _and_ stubbornness to pull it all off. Only in retrospect, the solution seems (kinda) obvious.<p><a href="http://www.amazon.com/More-Less-MacCready-Efficient-Flight/dp/1893554902/" rel="nofollow">http://www.amazon.com/More-Less-MacCready-Efficient-Flight/d...</a>
I enjoy my share of the new agey, platitude filled narratives that 37signals has been pushing out lately. But this one seems to have been written to push as many HN audience buttons as it possibly can. Sigh...