Facebook Now Using Artificial Intelligence To Prevent Suicides


You probably already knew of a suicide case where someone posted messages or videos on social networks indicating they were about to take their own lives. The social network giant Facebook wants to use these signals to identify users with this behaviour and thus prevent them from committing the act. To do so, the company is turning to artificial intelligence.

Facebook had to suicide prevention tools but needed something more proactive, so to speak. The new technology was being tested for some time in the United States, and this week it has been in use in other countries, with the exception of those in the European Union: the region's online privacy laws require access to sensitive personal information.

To function, the artificial intelligence system needs, in fact, to exploit the user's intimacy within the social network. The algorithm can analyze public posts, messages exchanged via Facebook Messenger, and live broadcasts to identify patterns that suggest suicidal behaviour.

Of course, the technology should speed up the reporting of suicide risks coming to Facebook, but the idea is to detect behaviours linked to the problem as soon as possible, even if no user has given the alert. Makes a lot of sense. Often a person who thinks about killing himself/herself makes subjective posts, which do not always make their intention clear.

One can, for example, say that no one cares about her or post sad pictures. They are common behaviours, which indicate moments of sadness, but not necessarily of suicide. So it is difficult to suspect something drastic.

Facebook's challenge is to use Artificial Intelligence to cross-link this information with other behaviours and trigger the alert. Any sign can be analyzed, even posts from contacts asking things "are you okay?" Or "Can I help?".

There are several approaches that Facebook adopts to prevent suicide. A user who reports a risk situation, for example, may receive instructions on how to act to help. The person with worrying behaviour may receive support notifications, with instructions such as talking to a friend or looking for a helpline.

But the main objective is to speed up the activation of emergency services or support entities. Facebook estimates that with the new tools, local authorities can be notified up to twice as fast.

Obviously, technology does not do it alone, after all, there may be false positives. The system detects a situation of risk or receives a complaint, triggers a review team and, if necessary, contacts the emergency services.

In this sense, the artificial intelligence algorithm fulfils the role of speeding up the work of the teams, for example, highlighting segments of videos that signal a worrying behaviour.

Even with all the technology, human awareness continues to be important in suicide prevention. By noting derogatory behaviour (such as bullying) or posts that encourage self-mutilation, for example, it is important to report to Facebook as soon as possible.

Likewise, it is important not to treat as "freshness" or "bullshit" the behaviour of acquaintances who indicate risk of suicide. Generally, people who take their own lives do not do so by the will to die, but by the perception that there is no other solution to the problems that afflict them. Under such circumstances, a simple conversation can help avoid the worst.

No comments

Powered by Blogger.