Facebook CEO Mark Zuckerberg, left, makes the keynote speech at F8, the Facebook's developer conference, Tuesday, April 30, 2019, in San Jose, Calif. (AP Photo/Tony Avelar )
Tony Avelar/AP
Facebook CEO Mark Zuckerberg, left, makes the keynote speech at F8, the Facebook's developer conference, Tuesday, April 30, 2019, in San Jose, Calif. (AP Photo/Tony Avelar )
Now playing
01:41
Facebook says that, unlike its past, its future is privacy
Shutterstock
Now playing
02:38
FTC antitrust complaint against Facebook dismissed by judge
Getty Images
Now playing
04:16
Emails show frustration over Facebook's handling of election lies
CNN/Getty Images
Now playing
06:01
Facebook VP: We are applying the most severe penalty we have on Trump
Getty Images
Now playing
01:41
Trump suspended from Facebook until 2023
Getty Images
Now playing
02:25
Facebook changes policy on Covid-19 origin claims
Now playing
02:37
Facebook Oversight Board: Indefinite suspension of Trump's account is 'not appropriate'
Facebook Trump board Welker Psaki White House response _00005421.png
Facebook Trump board Welker Psaki White House response _00005421.png
Now playing
01:56
White House responds to question about Facebook's decision on Trump
Now playing
04:11
Facebook decision 'is wait and see,' says former public policy director
ORLANDO, FLORIDA - FEBRUARY 28:  Former President Donald Trump addresses the Conservative Political Action Conference held in the Hyatt Regency on February 28, 2021 in Orlando, Florida. Begun in 1974, CPAC brings together conservative organizations, activists, and world leaders to discuss issues important to them. (Photo by Joe Raedle/Getty Images)
Joe Raedle/Getty Images
ORLANDO, FLORIDA - FEBRUARY 28: Former President Donald Trump addresses the Conservative Political Action Conference held in the Hyatt Regency on February 28, 2021 in Orlando, Florida. Begun in 1974, CPAC brings together conservative organizations, activists, and world leaders to discuss issues important to them. (Photo by Joe Raedle/Getty Images)
Now playing
01:32
CNN correspondent: This is a nightmare situation for Facebook
Facebook
Getty Images
Facebook
Now playing
03:14
More than 500 million Facebook users' personal data leaked online
Energy and Commerce Committee/YouTube
Now playing
02:50
US lawmakers question tech CEOs on misinformation
TOPSHOT - A demonstrator wearing a mask painted with the colours of the flag of East Turkestan and a hand bearing the colours of the Chinese flag attends a protest of supporters of the mostly Muslim Uighur minority and Turkish nationalists to denounce China's treatment of ethnic Uighur Muslims during a deadly riot in July 2009 in Urumqi, in front of the Chinese consulate in Istanbul, on July 5, 2018. - Nearly 200 people died during a series of violent riots that broke out on July 5, 2009 over several days in Urumqi, the capital city of the Xinjiang Uyghur Autonomous Region, in northwestern China, between Uyghurs and Han people. (Photo by OZAN KOSE / AFP)        (Photo credit should read OZAN KOSE/AFP/Getty Images)
OZAN KOSE/AFP/AFP/Getty Images
TOPSHOT - A demonstrator wearing a mask painted with the colours of the flag of East Turkestan and a hand bearing the colours of the Chinese flag attends a protest of supporters of the mostly Muslim Uighur minority and Turkish nationalists to denounce China's treatment of ethnic Uighur Muslims during a deadly riot in July 2009 in Urumqi, in front of the Chinese consulate in Istanbul, on July 5, 2018. - Nearly 200 people died during a series of violent riots that broke out on July 5, 2009 over several days in Urumqi, the capital city of the Xinjiang Uyghur Autonomous Region, in northwestern China, between Uyghurs and Han people. (Photo by OZAN KOSE / AFP) (Photo credit should read OZAN KOSE/AFP/Getty Images)
Now playing
01:16
Facebook: Chinese hackers targeted Uyghurs living in US
Chris, a Trump supporter, reacts to a fact check of a manipulated video shared by the Trump campaign.
CNN
Chris, a Trump supporter, reacts to a fact check of a manipulated video shared by the Trump campaign.
Now playing
03:58
What Trump supporters see on their Facebook feeds
Now playing
02:24
Under questioning, Zuckerberg admits Instagram was a 'competitor'
Facebook's CEO Mark Zuckerberg delivers his speech during the VivaTech (Viva Technology) trade fair in Paris, on May 24, 2018. (Photo by GERARD JULIEN / AFP)        (Photo credit should read GERARD JULIEN/AFP/Getty Images)
GERARD JULIEN/AFP/Getty Images
Facebook's CEO Mark Zuckerberg delivers his speech during the VivaTech (Viva Technology) trade fair in Paris, on May 24, 2018. (Photo by GERARD JULIEN / AFP) (Photo credit should read GERARD JULIEN/AFP/Getty Images)
Now playing
03:21
This is how Facebook kills its competition
San Jose CNN Business —  

On Monday, Facebook’s chief technology officer, Mike Schroepfer, tested my ability to tell the difference between broccoli and marijuana.

He showed me two pictures of green blobs and asked if they depicted the cruciferous vegetable or the mind-altering plant. I guessed both were cannabis; I was wrong. One, apparently, was an image of tempura broccoli.

Unlike me, Facebook’s content-filtering artificial intelligence technology can now determine which image is of food, and which is of marijuana, according to Schroepfer.

At Facebook’s annual F8 developer conference on Wednesday, Schroepfer detailed just how much AI has improved over time, and the ways the world’s largest social network uses the technology to, uh, weed out, posts and ads it doesn’t want users to see.

It remains tricky for AI to consistently identify objectionable content. Yet Facebook and other social networks increasingly rely on a combination of artificial intelligence and human moderators to police content posted by users. Facebook has been criticized for relying too much on human contractors, as well as for its AI not catching violent live-streams like the New Zealand mosque shooting in March.

“The New Zealand video was an awful example of a place where we need to do a lot better,” Schroepfer told CNN Business ahead of F8 over video chat.

Schroepfer said AI is now Facebook’s top method for finding bad content across most categories. And he explained that over the past four years, the company’s machine-learning capabilities have evolved from not being able to catch “the really obvious stuff,” to noticing content that isn’t easily recognizable as the kind of thing you wouldn’t want on Facebook.

Facebook CTO Mike Schroepfer explains at the F8 conference how AI is being used to find objectionable content.
CNN
Facebook CTO Mike Schroepfer explains at the F8 conference how AI is being used to find objectionable content.

Now, he said, Facebook’s automated systems are so capable at taking down images of ads with pictures of marijuana in them that people are posting ads for drugs with images of packaging, or of what appear to be Rice Krispies treats along with a short text description — both of which Facebook’s AI has also been able to detect.

“Today we’re catching stuff on a regular basis that I put it in front of regular people and they have to stare at it for a while and try to figure out why this is bad,” he said.

Yet while Schroepfer touted Facebook’s AI advances, he also acknowledged the recent shooting highlighted the challenges AI still faces in policing content.

When the suspected terrorist in New Zealand streamed live video to Facebook of a mass shooting, Facebook’s AI technology was of no help: the gruesome broadcast went on for at least 17 minutes until New Zealand police reported it to the social network. Recordings of the video and related posts about it rocketed across social media while companies tried to keep up.

Since then, Schroepfer said, Facebook has been studying what happened and how to find such violent videos faster. He thinks the company can improve as AI gets better at understanding why the content in the video is bad, or at picking up signals about the behavior of the person posting the video (like if they’d been flagged for issues in the past).

“We’re not where we’d like to be, so if people are criticizing us I get the criticism,” Schroepfer said. “I’m frustrated as anyone when we make a mistake or something happens.”

Many of Facebook’s AI advancements come from what’s known as supervised learning — an AI technique in which a computer is given labeled data (for instance, pictures of people that have each been labeled) so it knows what it needs to be on the lookout for.

But this can be labor-intensive for humans, making the technique ill-suited to tackling fast-moving content such as memes or election-related misinformation. To help find these quickly, Schroepfer said, Facebook has also made progress in what’s known as self-supervised learning, where a computer learns directly from data in an automated fashion.

He said Facebook recently rolled out a tool based on self-supervised learning to help with translating content related to the 2019 India election, which is currently underway. In March, the company deployed new software for spotting English-language hate speech in the US.

“Given the tenor of the conversation that’s happening and, honestly, the fact that we make mistakes on a daily basis, makes it hard for people to believe we’ve made a whole ton of progress,” he said. “I expect us to make a whole ton more progress over the coming years.”