Hollywood is no stranger to AI. Consider its role in movies such as Ex Machina and Blade Runner. Or in the popular HBO series Westworld.
But while AI may play a leading role in the entertainment industry’s depictions of the future on screen, it’s already starring in entertainment behind the scenes, thanks to Netflix.
Justin Basilico, research and engineering director at Netflix, stopped by NVIDIA’s AI Podcast to talk about how the streaming company applies deep learning to keep subscribers like us engaged.
“What we saw in … some of our own experiences internally at Netflix was that the typical deep learning approaches apply to the traditional recommendation system problem,” Basilico said in a conversation with AI Podcast host Noah Kravitz. “It wasn’t until recently in the past few years that really there’s kind of a paradigm shift where people are really starting to see the attraction in using [recommendations for deep learning].”
Basilico’s team works on the algorithms for personalizing the Netflix homepage, such as what artwork Netflix presents to depict each movie or series it recommends for each user.
So how does Netflix measure if their new algorithms are helping subscribers find content that they enjoy? According to Basilico, they track two user groups. One with the current service. The other with the new algorithm in play. They then analyze the long-term metrics.
“Because we’re a subscription company, our North Star is really whether people stay a Netflix subscriber over time, and we have a monthly subscription, so that means it typically takes a month or two to be able to measure that,” Basilico said. “But we see that as the ultimate signal that people see enough value to stay with Netflix. And if we can move that with the recommendation algorithm, we know that we are making people happier and making a better user experience and improving their satisfaction.”
In the years ahead, Basilico believes personalizing recommendations will play an even bigger role for Netflix and its subscribers.
“We’re also interested in thinking about problems that we have in terms of when we show recommendations to people and people click on it. Are they clicking on it because that’s what they really wanted to watch and it would be the best thing for them?” said Basilico. “Our goal really is to find the best content for each of our members that they’re really going to enjoy.”
If you’re interested in keeping up with Netflix research, check out the Netflix Tech Blog and their new research page.