Don’t Expect AI, Machine Learning to Replace Facebook’s 35,000 People Dedicated to Safety

Mark Zuckerberg said there are things that each do well

Humans and machines both contribute to Facebook's safety efforts AndreyPopov/iStock
Headshot of David Cohen

Machine or human? Human or machine? The debate resurfaced during Facebook’s third-quarter earnings call Wednesday.

Goldman Sachs analyst Heather Bellini referenced the 35,000 people Facebook has dedicated to safety and the company’s budget in this area during the call’s question-and-answer period.

Bellini asked, “How long does this have to be as resource-intensive from a people perspective? Or do you see the ability over time with technologies like artificial intelligence potentially helping you to get more leverage out of this existing budget? So maybe you don’t have to continue to ramp it as quickly as your user base and engagement continues to grow?”

Facebook CEO Mark Zuckerberg didn’t foresee dramatic changes on this front in the near future, saying, “In general, what we have to do is use computers and AI for what they’re good for, which is looking at a lot of content very quickly and making quick judgments. And we have teams of people doing what people are good for, which is making nuanced human judgments … And then there’s just so much content flowing through the system that we do need a lot of people looking at this. And I don’t think that that’s going to change anytime soon.”

Chief financial officer David Wehner chimed in, as well, saying, “When using AI and machine learning systems, it’s a real benefit to have actual people looking and tagging that content, classifying it, because it helps those machine learning algorithms be able to learn what they’re looking at.” David Cohen is editor of Adweek's Social Pro Daily.