Google Upgrades Its AI, Helping To Solve YouTube's Advertising Crisis

YouTube

Google is on the move to win back advertisers that are fleeing from YouTube amid their ads showing on extremists' videos. After two months defending itself, Google is making some changes to improve how contents are shown.

On April 3rd, 2017, a spokesperson for Google said that the company is working with "trusted" third-party firms to better identify offensive YouTube content that brands may not want to place ads next to. Google also said it is now relying on Artificial Intelligence to pinpoint contents that could worry advertisers.

Google has admitted that it had failed to spot extremists' contents as it revealed that it's teaching its AI to understand which videos are offensive. The tech giant also admitted that its computers were struggling to "understand context" and said it is using its "greatest machine learning abilities" to try to solve the issue.

To do this, Google deploys its AI algorithms to scan videos, analyzing them on a frame-by-frame basis. Google trains the AI to recognize videos by comparing those that are offensive with those that aren't, and comparing them with descriptions of contents.

The company hopes that it can teach its computers to understand the context of footage, so that videos that are easily recognized as offensive by humans, can be flagged automatically by its computers.

According to Philipp Schindler, Google's chief business officer, the company has been in "emergency mode" over this issue.

"We take this as seriously as we've ever taken a problem," he said. "Computers have a much harder time understanding context, and that's why we're actually using all of our latest and greatest machine learning abilities now to get a better feel for this."

Read: Big Advertisers Pulling Themselves From YouTube

YouTube filtering

YouTube’s vast video library makes it difficult to catch everything, at least for humans. It’s inevitable that some of the work will be delegated to AI algorithms.

Google’s Chief Business Officer, Philipp Schindler, told Bloomberg that human eyes contributed to the flagging, but that they couldn’t do everything. “The problem cannot be solved by humans and it shouldn’t be solved by humans.”

Over the past two weeks, Google has changed what types of videos can carry advertising, barring ads from appearing with hate speech or discriminatory content. In addition to that, Google is also simplifying the way advertisers can exclude specific sites, channels and videos across YouTube and Google’s display network.

It is allowing brands to fine-tune the types of content they want to avoid, such as "sexually suggestive" or "sensational/bizarre" videos.

The company also added more stringent safety standards by default, so an advertiser must choose to place ads next to more provocative content. Google created an expedited way to alert it when ads appear next to offensive content.

As a result, the company's AI flags five times as much objectionable content as usual in the last two weeks.

Google Ads

Google has been using algorithms to accomplish many tasks and automation. And over the years, it had trained computer systems to keep copyrighted materials and pornography off its YouTube service. But somehow, many of extremists, ISIS-related videos passed its filters.

After investigations, its engineers realized that their computer models had a blind spot: They did not understand context. This glitch in the company's automated process turned into a public-relations nightmare.

Many big brands, mainly from the U.S. and the UK were pulling their ads from YouTube. Companies like AT&T and Johnson & Johnson said they would pull their ads from YouTube, as well as Google’s display advertising business, until they could get assurances that such placement would not happen again.

More than 400 hours of videos are uploaded to YouTube every minute, and consumers watch more than one billion hours of videos every day. This puts YouTube at the dominant position on the internet, and an obvious means of advertising source for both Google and also video uploaders. As this is a significant problem for a multibillion-dollar company that still gets most of its revenue through advertising, the crisis opened Google to criticism that it was not doing enough to look out for advertisers, and it is putting hundreds of millions of dollars dent to Google's revenue in 2017.

While the upgrade can help Google in flagging inappropriate contents from ever showing on YouTube, however, the company also said that it will never be able to solve the problem completely since the number of uploaded videos on YouTube is staggering.

"No system can be 100 percent perfect," said Philipp Schindler. "But we're working as hard as we can to make it as safe as possible."