Facebook makes its addictive AI platform open source

Facebook is addictive— that’s a fact, and it’s likely the key reason why the Cambridge Analytica scandal didn’t sink the platform this year. Instead, it continues to bob safely on choppy waters.

Besides leveraging our inherent, voyeuristic tendencies and natural desire to be connected, the social media platform employs sophisticated AI designed to maximize its users’ screen time, and keep them coming back again, and again.

That ability comes courtesy of a suite of coding tools, Horizon— a Reinforcement Learning (RL) platform, which allows for the training of algorithms for the continuous optimization of personally-relevant push notification which it predicts will result in the best click-through rates and “meaningful interactions”. The model receives a reward when someone engages on a post they would otherwise miss and a penalty for sending the notification.

That’s just one of many uses, though. “The platform is already providing performance benefits at Facebook, including delivering more relevant notifications, optimizing streaming video bit rates, and improving M suggestions in Messenger,” according to Horizon’s lead engineer, Jason Gauci.

“But Horizon’s open design and toolset also have the potential to benefit others in the field, particularly companies and research teams interested in using applied RL to learn policies from large amounts of information.”

Yesterday (November 1), Facebook announced that Horizon would be made open source, meaning developers can tap into these powerful capabilities to maximize engagement with their users. According to Gauci, the technology is available to everyone from a “hobbyist or high school student” or to be run on “thousands of machines in the cloud”.

“Horizon is part of our commitment to the open development of AI — it was an internal Facebook platform that we’re now open-sourcing, and it works with tools that we have already made available to the community, including PyTorch 1.0 and Caffe2,” said Gauci.

“In addition to making an impact on new applications, Horizon could transform the way engineers and ML [machine learning] models work together. Instead of driving models by writing rules that are difficult to interpret and maintain, we envision a two-way conversation between engineers and the models they work with, in which engineers can specify their high-level goals, and work in tandem with machine learning to realize these goals and adapt them to an ever-changing decision landscape.”

As is usually the case behind decisions to ‘go open’, Facebook is looking to encourage the development of its cutting-edge technology through experimentation and help in recruitment, with applicants already familiar with the tech.

As noted by Engadget, going public with Horizon is also a strong PR move in showing off the platform’s level of sophistication against rivals like Google and Microsoft.