Facebook may soon rely to ‘Artificial Intelligence to identify – and eliminate – the false news that are daily published within the social network.
The news was provided directly by the Head of the AI Yann LeCun who said that using artificial intelligence could also be possible to “ eradicate ” other types of incorrect content, including violent videos. The same LeCun he admitted that there may be a dangerous trade-off between “ complaint ” and “ filtration ” which could undermine the credibility of the same social network.
“What is the trade-off between filter and censor? Freedom of experience and of decency? “, He wondered LeCun during a round table. “The technology already exists or can be developed. But then the question is how it can make sense to resort to this. And this is not in my field. “
E ‘for some time that Facebook is working intensively on artificial intelligence to provide instruments to ensure the best user experience for its users, as well as to monitor the site by removing any content deemed dangerous to public security or discriminatory and to reduce as much as possible the presence of clickbait.
Use artificial intelligence to automatically censor certain types of content, however, could also lead to an incorrect result, for example going to delete posts that a manual check not erase .
What is certain is that Facebook is trying to identify the best solution to eliminate false news, which is why, in addition to the development of artificial intelligence, Zuckerberg recently hired hundreds of people called to monitor “live” content published . It is a solution-buffer, not able to identify all cases reprehensible, but it testifies to the efforts that the company is doing to improve its service.
The introduction of new features such as live video has also made it more difficult to monitor the content: for these cases are required ultra-fast calculations of algorithms that are able to block the vision within a few seconds.
Artificial intelligence is an integral part of Facebook: behind the scenes you are already using this technology to show in each of the 1.79 billion users at least once a month to access the service news that most interest.
Investments have started three years ago, and now the development team headed by Yann LeCun are faced with a new – and more difficult – challenge that combines technology with ethical issues, with trying to “ avoid doing damage and systematic biases . “