Google CEO Sundar Pichai says he’s not sure YouTube will ever be able to fully eliminate hateful content on the video-sharing site.
YouTube has increasingly relied upon a mix of humans and technology to take down videos that violate its rules, but it may never get rid of this content completely, Pichai admitted in a new interview with CNN.
It comes as the Google-owned platform, which counts two billion-plus users worldwide, has come under fire for its failure to remove content promoting hate speech, conspiracy theories and violence.
Scroll down for video
YouTube relies upon a mix of humans and AI to take down videos that violate its rules, but it may never get rid of this content entirely, according to Google CEO Sundar Pichai (pictured)
YouTube is working hard to get rid of bad content, but given its ever-increasing scale, this problem has become more and more challenging, the Google CEO explained.
‘It’s one of those things where let’s say we’re getting it right over 99 percent of the time, you’ll still be able to find examples,’ Pichai told CNN.
‘Our goal is to take that to a very, very small percentage. Well below one percent.
‘Any large scale systems, it’s tough. Think about credit card transactions, there’s still going to be some fraud in the system. So anything you run at that scale, you have to think about percentages,’ he added.
YouTube continuously evolves its policies around these kinds of content in order to respond to changes in user activity or new issues that arise, he said.
For example, earlier this month, YouTube announced it would ban users from posting videos that promote violence and extremism, such as Nazi glorification and white supremacy, as well as hoax videos that try to debunk known tragedies, like the Sandy Hook shooting and the Holocaust.
In an interview with CNN, Pichai explained that he wishes Google had addressed some of YouTube’s issues earlier, such as the spread of hoax videos denying the Sandy Hook shooting
Earlier this month, YouTube announced it would ban users from posting videos that promote violence and extremism, as well as hoax videos that try to debunk known tragedies
Despite the new policy, YouTube has received criticism for its slow response to many of its most pressing issues, such as the presence of hoax videos that deny the existence of the Sandy Hook shooting or the Parkland shooting.
‘We wish we had gotten to the problems sooner than we did,’ Pichai told CNN.
‘I think we became aware collectively of some of the pitfalls here and since then, we’ve been working hard. We’ve changed some of our priorities.’
Along with banning some hateful content, YouTube has also ramped up its focus on ‘borderline content.’
This includes content that doesn’t necessarily violate its policies, but could be used to spread ‘harmful misinformation.’
Some have argued that government regulators should step in to better manage Silicon Valley’s issues around hate speech, privacy and potential anti-competitive practices.
Pichai pushed back against this idea, telling CNN: ‘Scale does offer many benefits, it’s important to understand that. Big companies are what are investing in technologies like AI the most.’
If tech giants’ scale were to be cut back, it could create an opening for other countries to become dominant players, he added.
‘There are many countries around the world which aspire to be the next Silicon Valley, and they are supporting their companies too,’ Pichai said. ‘So we have to balance both.’
In a separate interview with Axios, Pichai said YouTube is applying some of the tactics used in Google Search to rank content by quality in an effort to stamp out borderline material.
‘We are bringing that same notion and approach to YouTube so that we can rank higher quality stuff better and really prevent borderline content,’ Pichai told Axios.
‘Content which doesn’t exactly violate policies, which need to be removed, but which can still cause harm.’
In order to do so more effectively, Pichai said Google and the industry as a whole will have to answer some tough questions.
‘It’s a hard computer science problem, it’s also a hard societal problem because we need better frameworks around what is hate speech, what’s not, and how do we as a company make those decisions at scale and get it right,’ Pichai told Axios.
‘We aren’t quite where we want to be,’ he added.
