• Dr. Timothy Smith

Could a Robot Be a Good Therapist?: Artificial Intelligence and Mental Health


Photo Source: Pixabay


Mental health issues affect millions of people around the world, but the number of psychologists and psychiatrists remains woefully lacking. According to the National Institutes of Mental Health (NIMH), nearly one in five or 51.5 million Americans over the age of 18 live with some sort of mental illness. (nimh.nih.gov) With the rise of artificial intelligence, it lends one to wonder if it can help better diagnosis and treat mental health issues.


Mental disorders, according to the NIMH, vary from mild to severe and cover a variety of conditions, including (but not limited to) depression, post-traumatic stress disorder (PTSD), eating disorders, autism spectrum disorder, and ADHD. The same NIMH data predicts that only half of people with mental health issues ever seek treatment. Moreover, not everyone has equal access to mental health professionals. In a 2015 report from the American Psychological Association titled, “Demographics of the U.S. Psychology Workforce,” the authors demonstrate that over one third (34.7%) of all the active psychologists in the US are concentrated in only four states: California, New York, Pennsylvania, and Massachusetts. According to the same report, the state of Mississippi has only 160 practicing psychotherapists, which amounts to one therapist for every 18,600 citizens in contrast to California, which has one psychotherapist for every 3,600 citizens. (apa.org) For the most part, psychotherapy occurs either one on one between a patient and a therapist or in small group settings. In either case, human therapists cannot be in more than one place at a time, severely limiting the scalability of mental health care.


Scaling mental healthcare to reach more people continues to challenge the field. In the past, print media such books or pamphlets and visual media such as videos have served to help people wanting help but not interested, ready, or able to engage mental health professional. Today, a number of institutions and startup companies have used artificial intelligence to develop mental health applications such as chatbots to serve as an alternative option to working with a human. The market continues to swell with the new applications to help with everything from stress management with guided meditation such as Headspace.com to cognitive behavioral therapy to aid in weight loss such as Noom.com. The University of Southern California Institute for Creative Technologies (ICT) in conjunction with the Defense Advanced Research Projects Agency (DARPA) are working to build an AI driven application called Simsensei. The application combines language processing to listen to and speak with a patient as well as emotion detection with computer vision to identify distress, depression, anxiety, and PTSD. Simsensei, nicknamed “Ellie,” began in 2011 and continues to evolve today (ict.usc.edu) Using emotion detection and a simulated therapist, Simsensei provides an alternative, scalable option to patients to get mental healthcare and it helps doctors to monitor patients for changes in mental health over time without the need for in person visits.


Mental health issues affect many people around the world. In the US alone, nearly 20% of adult over 18 years of age will suffer some sort of mental health issue in their lives such as depression, ADHD, PTSD, and others. In spite of the large number of people affected, projections indicate that only half of afflicted people will seek any form of treatment. Some will not seek treatment due to the lingering stigma associated mental health and others simply do not have access or cannot afford it. Certain states such as California and New York have a much higher number of the active psychotherapists than other states such as Mississippi, Nebraska, or the Dakotas. Because of the one-on-one nature of therapy, scaling psychotherapy poses a continuing challenge. However, artificial intelligence driven robotic therapy now offers broadly accessible alternatives that anyone with access to the internet can use. Companies and governmental agencies continue to fill the market with various application such as Headspace, Noom, or Ellie. Such applications continue to proliferate, but how well they work still needs further research. In a recent paper titled,” Changes to the Psychiatric Chatbot Landscape: A Systematic Review of Conversational Agents in Serious Mental Illness” by researchers at Harvard Medical School, the authors reviewed all the literature on the effectiveness of currently available psychiatric chatbots. They looked at 247 studies and found only seven of them were good enough to trust. They concluded that only a few chatbots are meeting expectations and need more research to prove they work. The promise of AI to help reach the millions of people with needed help remains possible, but it will take more time and research before the robot therapist becomes a reality.



Dr. Smith’s career in scientific and information research spans the areas of bioinformatics, artificial intelligence, toxicology, and chemistry. He has published a number of peer-reviewed scientific papers. He has worked over the past seventeen years developing advanced analytics, machine learning, and knowledge management tools to enable research and support high level decision making. Tim completed his Ph.D. in Toxicology at Cornell University and a Bachelor of Science in chemistry from the University of Washington.

You can buy his book on Amazon in paperback here and in kindle format here.