The New York Time technology blog, Bits, has an update on the Netflix prize. No team has made the Netflix recommendations engine 10% more accurate, which is the number required to claim the million dollar prize. However, Netflix has awarded $50,000 to the team that has made the most progress. The group that won the prize has spent around 2000 hours on the project so far, and is made up of researches from AT&T Labs. There’s no word on whether the team worked on the prize during work hours or in their off time. Either way, their compensation comes down to about $25 per hour, which may be a lot to pay for lawn care but is an incredible deal for the services of ordinarily well-compensated computer scientists. All of the other teams that submitted entries weren’t compensated at all.
It looks to me like prizes can be a great way of funding research and development so long as the problem to be solved is interesting to geeks. I suspect that this is something Google already knows.