Forget algorithms. Send in the critics.

I don't know about you, but I've trained my eyes to ignore the "personal" recommendations offered on sites like Netflix or Amazon. They just leave me feeling so, so... misunderstood. My friends at Amazon still insist that I buy the latest military history, because I once, in a fit of gift guilt, sent a tome about West Point to a distant relative. And the helpful fellows at Netflix know that I will ♥ "Dora Saves the Mermaids." Don't ask.

Anyway, the current Wired has an intriguing piece about the mechanics of recommendation software. It seems Netflix is sponsoring a contest that will award $1 million to the person or team that invents a better movie-trap. The Million Dollar Question is this: How do you write an algorithm that defines taste? The usual suspects -- data miners from prominent universities and tech companies -- have lined up to meet the challenge. This has been going on for more than a year, with infinitesimal improvements in performance and plenty of amiable cooperation among the competing teams.

Enter a contestant known simply as "Just a Guy in a Garage,'' a worthy competitor who came out of nowhere and got terrific results. Wired's intrepid reporter, Jordan Ellenberg, tracked "Just a Guy" down, and it turns out he's an unemployed psychologist in England who takes a slightly different approach than the math-heads. Ingenious programming, he implies, is essential, but it isn't good enough unless it takes human behavior into account. He's trying to figure out a way to bring the math and the mind together, and in the process, take home the chunk of change.

Here's one possible solution. "Just a Guy" needs to figure out a way to match the customer data to a database of critics' opinions. Think about it. Don't we all have our favorite critics, those writers who almost always nail our exact thoughts about a given film, play, or other performance? If the technical wizards can write a formula to match the user to the critic, they then have a surefire way to improve the accuracy of their predictions. I'd much rather have my movie suggestions tied to the taste of say, Anthony Lane, than to a string of ones and zeros that are as emotionally remote as the solution to the Towers of Hanoi. Algorithms can't think, breathe, or spend a lifetime sitting in dark theaters. That's why we'll always need critics. Now if only I could do the math, I'd be able to retire.