Blog Stats
  • Posts - 102
  • Articles - 32
  • Comments - 212
  • Trackbacks - 16

 

The Netflix prize

The past two weeks, I've been in Boston, Denver, Las Vegas, Minneapolis, and Providence. Needless to say, it's been a bit tiring. Fortunately, the next several weeks I get to work out of my office and then I'm back on the road again for a bit. I have an article on hiring for single founder companies that I'm nearly finished with. Also due up is an article on how to find clients if you've decided to build a consulting company to help fund your software company.

Of course, if you don't feel like spending all that time doing consulting work and think you're the best hotshot programmer since Hugh Jackman in the movie Swordfish (go ahead... spin your chair when programming. All the cool kids are doing it and it really does make you a better programmer...), you can skip that part and try to snake yourself a cool million dollars by winning the Netflix Prize.

"What's that?" you ask?!?!

Let me enlighten you my friend. Netflix is holding a contest where they are providing a set of non-personally identifiable data about some of their customers and their viewing habits. If you can write an algorithm that will predict what people might like to watch that is 10% better than their current algorithm, they'll pay you $1 million dollars. (Insert Dr. Evil hand gestures and maniacal laugh here).

The contest has been going on for a while now, but I really haven't heard too many people clamoring about it on the websites I tend to frequent. Less than 11,000 entries have been submitted so it's not as if you have a snowball's chance in hell to win. You just have to come up with an algorithm that works better than 11,000 other algorithms.

While this isn't something that I would participate in at this stage of my life, I suspect that there will be more contests like this in the future from other companies to harness the collective talents of people from all over the world. For Netflix, this type of algorithm development is a bargain. They're paying for results, not for research and development that might or might not pan out. It's similar to buying a startup company, but they have already specified the output they want, and if it doesn't happen, then Netflix doesn't pay a dime.

Pretty ingenious if you ask me. Kudos to Netflix.

Feedback

No comments posted yet.


Post a comment





 

Please add 3 and 1 and type the answer here:

 

 

Copyright © Mike Taber