Can AI replace humans in psychology?

Why are so much time, energy, and financial resources being poured into e-health? Because mental distress, particularly among young people, is a global pandemic.

The latest World Health Organization study shows that one in five teenagers experiences mental distress, and research confirms that some 90% of young adults ages 18-29 in the United States utilize social media, preferring text to phone calls. A study done this year in Hong Kong found that 30% of students who committed suicide shared it online first but no one detected those calls for help.

During COVID-19, there has been a dramatic increase in mental distress. Recently, there was a tragic case of a 12-year-old boy who jumped out of his bedroom window while his mother and sister were in the other room.

Mental health helplines, which are set up to work with people suffering from mental distress, are inundated with calls for help. A recent study in the European Union shows that of the nearly 2,000,000 appeals for help, approximately half – yes, 1,000,000 – went unanswered. The professional helplines, whether run by the government or by non-profits, simply cannot keep up with the workload.

The cost of failing in mental health is great, not only in the costs associated with the interventions required when the situation is not detected in time, but also in the psychological or physical damage done to the people in distress and those around them. 

This week is “Suicide Awareness Week” in Israel. To date, very little has been done to help curb the number of suicides. There have been studies done, research conducted, many recommendations put forth and meetings held, but when it comes to tachlis – practical matters – almost nothing.

I am the creator of MoodKnight, an artificial intelligence based software solution that has the proven ability to detect and prioritize mental distress situations online, in real time. The MoodKnight software has consistently scored in the range of 92% accuracy in a review of millions of posts from various social media websites.

MoodKnight understands human language and particularly texting, slang, emoji-speak, and more. A digital risk assessment is created and the post is categorized as mild, moderate or high risk. High-risk situations are immediately referred to the 105 Mental Distress Helpline or to a nonprofit that provides anonymous, immediate, free online help.

In most cases, posts with risk levels in the mild-to-moderate range can be helped from within the network, and dramatically reduce distress by use of peer-to-peer support and trained volunteers.

This year, MoodKnight received a second round of grants from the Israel Innovation Authority and has become part of the Digital Israel initiative. Over the last few weeks, we presented during the Ministry of Health’s Health IL 2020 Innovation Week, and demonstrated MoodKnight’s software at the Knesset session on Science and Technology: Technological Resources for Protecting Children and Youth Online. Eti Shmueli, director of the Communications Engineering Division of the Ministry of Communications, was present and her response to the technology was “We need this technology!”

We cannot replace the human touch that is necessary when dealing with mental health issues, but we can greatly assist mental health professionals in doing their job more efficiently, effectively, and successfully.