Facebook Inc (NASDAQ:FB) is spending massive amounts on research and opening labs around the world with one goal in mind. The firm wants to see everyone, and it wants to recognize them. Its new feature “Moments” shows the power of that goal and advertisers are hungry to exploit the potential of the firm’s ability to know what it’s looking at.
Facebook is making great strides in its development of AI tech, and the image recognition of photos is just the first step. The sheer amount of money and talent that the firm is pouring into research in the area means that it will likely set a course for the its future.
Facebook will recognize your friends
“Moments” will allow a Facebook user to make a group of friends that they want include in a photo album. Facebook’s AI will then look for photos with them in it and group them together. This will allow Facebook’s users to make albums of events that their friends were at without needing to group them manually.
The firm’s facial recognition algorithm, driven by Mark Zuckerberg’s obsession with AI, boasts accuracy of 98% and will be able to see people in photos even if their faces are partially obscured.
Facebook says that the system is not able to recognize people, it’s only able to see if two people are the same. That’s a legal more than a technical hurdle, however. When the firm first brought its face recognizing app out to the public it faced a huge backlash.
The ability to recognize people in photos is the first step in Facebook’s campaign to teach machines to understand the world without context from humans. By learning patterns, machines are able to quickly see the mistakes they make and get better at their task.
Making machines see the world at Facebook
Building machines that learn better will enable Facebook to make its network understand human behaviors and give advertisers access to much more powerful tools to reach the people who are most affected by their ads.
For now, however, Facebook is staying far away from the kind of trouble bringing in a system like that could cause. Yann LeCun, head of the Facebook AI program, made that very clear when speaking to Fortune about the new system.
“We don’t care who is in the picture. We would use other types of image recognition and train them differently to say that this looks embarrassing and then tap you on the shoulder to make sure you want to post this publicly,” he told the outlet.
Facebook doesn’t offer the photo-tagging service in the EU or Canada because those places have stricter privacy laws than the USA.
LeCun may not care about who is in the photos, but ad-makers and sellers sure will. He says that “Anything that can happen in the digital world we want to understand the context.”
If Facebook understands it, it’s certain that its ad team will figure out a way to sell it to the firm’s ad partners. Facebook shares have risen by more than 25% in the last year alone, and the only way the firm can justify that kind of value is by boosting the amount that is spent on ads on the site.
George Zarkadakis, author and AI expert, says that “It’s really important that we take AI seriously. It will lead to the fourth industrial revolution and will change the world in ways we cannot predict now.” Facebook is way ahead of him.