Asking an AI chatbot to give a rundown on Napoleonic wars is fine. But using a chatbot service for a therapy session?
Even ChatGPT suggests going to a traditional mental health practitioner when you pour your heart out to the AI – perhaps because the most important element of therapy is the client-therapist relationship.
But one of the biggest hurdles people face in their quest to get the best mental health service is the lack of abled practitioners. The demand is high, and good psychologists often don’t have the capacity to see new patients.
There was a survey conducted last year by the American Psychological Association which found that six in 10 practitioners in the U.S. don’t have openings for new patients. And on average, psychologists reported being contacted by over 15 potential new patients a week – a trend that became more prominent during the Covid-19 pandemic.
Digital mental health applications may offer a viable solution to this problem, but the big question remains.
Would you open up to an AI chatbot therapist?
Researchers from the United States have built an artificial intelligence (AI) voice assistant app called Lumen, which delivers a form of psychotherapy that can help patients with mild depression and anxiety.
For their study, Lumen was integrated within the Alexa app on an iPad provided to 63 participants. Two-thirds of the patients used Lumen for eight problem-solving therapy (PST) sessions, while the remaining participants were part of a control group receiving no intervention.
PST is a brief psychosocial treatment for patients experiencing depression and distress related to inefficient problem-solving skills.
The results of the study revealed that the Lumen group showed decreased levels of depression, psychological distress, and anxiety in comparison with the control group. The researchers said that the Lumen group also showed improvements in problem-solving skills.
“This kind of technology may serve as a bridge. It’s not meant to be a replacement for traditional therapy, but it may be an important stop-gap before somebody can seek treatment,” said Dr. Olusola A. Ajilore, a professor of psychiatry at the University of Illinois Chicago and co-first author of the paper.
The team thinks that their pilot study could be a step in the direction of filling a desperate need in the mental health system. The research group will now conduct a larger trial to compare the use of Lumen with both a control group and patients receiving human-coached PST.
The team will also consider integrating the dramatic progress recently made in AI language models, such as ChatGPT, for future iterations of Lumen or other digital interventions. But with all such AI uses in medicine and psychiatry, the hype must be confirmed with careful research, said a statement.
“It’s about changing the way people think about problems and how to address them, and not being emotionally overwhelmed,” said study senior author Dr. Jun Ma. “It’s a pragmatic and patient-driven behavior therapy that’s well established, which makes it a good fit for delivery using voice-based technology.”
Mental health is just like physical health; it requires upkeep
An estimated 5% of the population, roughly 280 million people globally, have depression. But, across a range of social and demographic groups, there are many barriers to accessing mental healthcare, such as socio-economic situations, financial barriers, lack of mental health education, ethnicity, gender, and age, among others.
While AI chatbots for therapy are an underexplored area with considerable potential for behavioral counseling and the promotion of emotional well-being, there are some major problems plaguing the technology.
Chatbot hallucinations: Sometimes, these chatbots spout out false information and, by doing so, may cause distress to their user. For someone undergoing therapy, this is a big no-no. Simply put, there is no room for spouting out falshoods as far as therapy is concerned, and developers have to be absolutely sure about their product if it is to be rolled out for public consumption.
The study was published in the journal Nature.