The recommendation system for popular movie-streaming website Netflix completely understands Gregory Spiegel ’13. From last summer’s sleeper hits to droll BBC programming, every single suggestion the site has made for him over the past four months has felt handpicked for Spiegel himself.
The algorithm, which matches users with content based on stated preference and previously viewed material, predicts his taste with an unheard-of 100 percent efficacy, allowing the student to fill his day with hour upon hour of on-demand video. “What can I say?” Spiegel replied when reached for comment. “Netflix gets me. Totally.”
Spiegel’s taste profile understands him and perfectly reproduces that understanding in the form of suggested media. Spiegel confirms that it is almost like the site knows him personally, lovingly compiling a unique cyber-library especially for him. Whether the biology concentrator cares to watch something Gritty, Sentimental, or Inspiring, the web service knows exactly what to choose and when to choose it.
Not once has Spiegel been let down. “I always joke that Netflix knows what I want before I do, knows what I need before I need it,” he said. And Netflix is prepared to meet that need, whether that means recommending “Twin Peaks” or 2003’s “Lost in Translation.” With every title watched and rated, the Netflix recommendation algorithm grows in strength.
Netflix has completely freed Spiegel from the burden of choice, piping top-quality content into his 17-inch laptop computer monitor 24 hours a day. The website commits his every instant entertainment choice to flawless memory. Spiegel in turn tells the site what he is interested in and dutifully assembles his Instant Queue, content to know that the website knows him in his totality. Netflix now knows over 26 million subscribers worldwide. Netflix is learning.