January 5, 2007
It’s been a while since Netflix offered a $1M prize to anyone that could build a better movie recommendation engine. The result thus far shows lots of teams have beaten Netflix’s system by at least 4%, with some fast approaching the 10% improvement mark to win the prize. The question, however, is why has it been so easy for so many people to beat Netflix? One would think they could have hired a team from a top-notch AI group for $1M to build a brilliant prototype. Instead, they rolled their own and pompously declared they’d “be seriously in the running for a Progress Prize, if we were eligible”. Big deal… everyone has topped the Progress Prize mark. At this point, Blockbuster should offer the winning team an outrageous sum of money to develop their site. In fact, companies should be bidding large sums of money for all these programmers since they’ve clearly demonstrated how good they are. In this case, recruiting good people is easy.