Netflix, Inc. (NASDAQ:NFLX) puts a lot of effort into figuring out what people might want to watch, but now it’s upping the ante by using the latest advances in Deep Learning to create an artificial brain that can understand what connects the movies you’ve enjoyed and what movies you have to see.

Netflix

Deep learning is a field of artificial intelligence focused on creating neural networks that simulate the brain in a very direct way. Just like the brain can increase or decrease the strength of the connections between neurons, deep learning creates a neural network consisting of nodes. The connections between the nodes can be adjusted by the neural network to become better and better at solving real world problems, a process called training.

Deep learning has only recently become practical

The basic idea isn’t new, but putting it into action, even on a relatively small scale, takes so much computational power that it has been impractical until recently.

“Many researchers have pointed out that most of the algorithmic techniques used in the trendy Deep Learning approaches have been known and available for some time. Much of the more recent innovation in this area has been around making these techniques feasible for real-world applications,” says a recent post on the Netflix tech blog.

What’s different now is that Andrew Ng and his team of researchers have found more efficient implementations that run on GPUs (graphics processing units) instead of CPUs, and which are able to train a large neural network faster than any previous protocols. This has led to a resurgence in deep learning as other companies, including Google and Facebook, to hire deep learning researchers to improve their own products as well.

Netflix will lean on the AWS cloud

But to save even more money, Netflix, Inc. (NASDAQ:NFLX) has decided there’s no need to build all the computing infrastructure itself, and will instead rely on Amazon Web Services to supply the bulk of the necessary computing power.

“We wanted to use a reasonable number of machines to implement a powerful machine learning solution using a Neural Network approach. We also wanted to avoid needing special machines in a dedicated data center and instead leverage the full, on-demand computing power we can obtain from AWS,” says the post.