November 11, 2016
The media industry’s interest in artificial intelligence goes much deeper than simply portraying its implications in movies such as “Her” or “Ex Machina.” Recommendations and push notifications are just two examples of how media uses AI. YouTube has evolved its use of machine learning algorithms to improve its content recommendations. In the early days, the site used “collaborative filtering” to feed videos to viewers. Now the company uses much more complex models based on deep learning powered by neural networks.
Variety quotes YouTube engineering VP Cristos Goodrow as saying that, “We believe that for every human being on the planet, there are already 100 hours of videos on YouTube.” Twenty years ago, Goodrow worked on creating neural networks for “big enterprise” clients, but the technology wasn’t reliable.
“Today, they suddenly work really well,” he said, noting that, “the key difference is data.”
YouTube now has “hundreds of billions of data points of viewing behavior,” that the deep learning algorithms can take advantage of. Right now, Goodrow is using AI to dig deeper into figuring out which videos customers “truly love.”
Google chief exec Sundar Pichai has said that the company’s “shift to AI is as fundamental as the invention of the Web or the smartphone.”
“We are at a seminal moment in computing,” he said. “We are evolving from a mobile-first to an AI-first world.” Pichai is not alone in that belief. At AI startup Vidora — which is working with News Corp, Yahoo Japan and Walmart’s Vudu video service — chief exec Alex Holub predicts, “In the next three to five years, every business will have an AI engine at its center.”
Vidora offers technology for companies to “customize their communications with consumers to reduce churn, be it on their websites, in apps, or via mobile push notifications,” and to learn “not only which videos to recommend, but also how to market each video, and even find the best price and value proposition for any given customer.”
Due to the massive amounts of online data now available, Holub believes that AI systems will soon be able to “develop complex strategies.”