Facebook outlines user experience objectives for AI

FacebookFacebook has outlined its ambitions for artificial intelligence and machine learning capabilities to enhance user experience.

The recent introduction of bots through the Facebook platform is one of first steps on the journey to artificial intelligence, which the team believe can evolve into an AI platform which can learn and automate specific activities. The Facebook team ultimately want to build computer services that have better perception than people, whether this is predicting what content would be relevant to a user or products would be of interest, which it believes is possible within the next 5-10 years.

“We’re focused not on what Facebook is, but on what it can be and on what it needs to be, and that means doing bold things,” said Facebook Founder Mark Zuckerberg. “A lot of what we’re building today in areas like connectivity, artificial intelligence, and virtual and augmented reality may not pay off for years, but they’re important to our mission of connecting the world. And I’m committed to seeing this mission through and to leading Facebook there over the long term.”

While true artificial intelligence could be perceived as a long-term ambition of the tech industry, Facebook has incorporated various AI and machine learning capabilities into its services over recent months. The Moments app is using face recognition to help users share pictures with friends, while also using AI to drive relevant content through a user’s news feed and filter spam. One of the more advanced applications of the technology is helping blind people comprehend what is in a photo by reading explanations of them aloud.

In terms of long-term ambitions for AI and machine learning capabilities will be to enhance the user experience and continue to drive more relevant content through their Facebook accounts. The team believe the future of AI will be able to understand the content of articles or videos in a more complex manner, linking the specific content with a user’s defined interests and previous use of the platform.

Currently, AI can potentially list what content is within an article, picture or video, but it doesn’t fundamentally understand what this content is, and thus cannot draw conclusions as to which users it would be relevant for. This intuition and perception would appear to be the next step in Facebook’s AI journey.

“One obvious thing I think over time is if you just look at the way that we rank News Feed, today we use some basic signals like who you’re friends with and what pages you like as some of the most important things for figuring out what – out of all of the millions and millions of pieces of content that are on Facebook, what we’re going to show and what are going to be the most interesting things to you,” said Zuckerberg.

“That’s because today our systems can’t actually understand what the content means. We don’t actually look at the photo and deeply understand what’s in it or look at the videos and understand what’s in it or read the links that people share and understand what’s in them, but in the future we’ll be able to, I think in a five or 10-year period.”

From a financial perspective, revenues for the quarter grew by 52% year on year to $5.4 billion, and advertising revenue grew by 57% to $5.2 billion. Mobile advertising revenue reached $4.2 billion, up 75% year over year, and is now approximately 82% of total advertising revenues for the business. The company now claims to have 3 million active advertisers on Facebook and over 200,000 on Instagram.