There was once a time when it was cool and profitable to build sparkling new software algorithms. The classic example of this was a bespoke search engine that two Stanford PhD students created, Google. In 2004, Google IPO'd for a record $24.6Bn. Groupon is rumored to be readying to IPO for $25Bn, setting a new record. With little algorithmic wizardry behind it, Groupon perhaps proves that technologists the caliber of rocket scientists are no longer necessary to create hit startups.
Fundamental problems in using the Internet, such as search and payments, are perhaps 'solved' to at least a 'good enough' level. These were the the type of problems best suited for technologists. Groupon proves we now need people, and startups, that can solve higher level problems. Perhaps the type of rocket scientists we now need are those that truly understand the problems people are trying, yet struggling, to solve through the Internet - such as get group discounts to restaurants and other merchants.