phonlamaiphoto - stock.adobe.com
As the new coronavirus spreads around the world, social media companies including YouTube and Twitter are ramping up automated content curation to fill in for employees who have been forced to work from home.
Automated content curation
YouTube and Twitter have long used a mix of machine learning systems and human employees to detect and remove content that violates their policies. Historically, the companies have used machine learning systems to identify potentially harmful content, which would then get sent to a human employee for review.
Both Twitter and YouTube noted their systems are prone to errors, and cautioned users that they may see more content accidentally removed than normal.
The systems are not fully ready to work on their own, according to Alan Pelz-Sharpe, founder of market advisory and research firm Deep Analysis.
"They are not ready, but the necessity to do something quickly surpasses the need to wait until they are," Pelz-Sharpe said.
"At this point, these systems should have been ready, but the cost and complexity and the lack of a sense of urgency means they are not," he continued. "The situation over the last couple of weeks has changed, and the urgency is there now."
Alan Pelz-SharpeFounder, Deep Analysis
The AI systems will learn, however, Pelz-Sharpe said, and in the long term they will improve on their own.
Twitter noted in its blog post that, due to the increased use of its AI systems, it will not permanently suspend any accounts based solely on its automated content curation system.
Instead, it said, it will "continue to look for opportunities to build in human review checks where they will be most impactful."
Businesses of all types will likely find ways to use AI more as their workforces are disrupted and displaced, potentially helping to further drive AI in the enterprise.
"The most obvious way is through more and more automation of jobs. Smaller tasks that require key entry will be automated and at unprecedented scale," Pelz-Sharpe said.
"Some of that will be done now where possible, but post-apocalypse this will become a major driver for AI -- the automation of the myriad of relatively simple tasks that today are ignored in favor of grandiose AI projects. It's a fundamental shift in focus for AI," he said.