AI, NLG, and Machine Learning
Three Bots That Can Help You Deal with Mental Health Issues
As the online therapy trend keeps growing, therapists may struggle in keeping up with patient demand, leaving some people in need of alternative forms of therapy. Artificial intelligence (AI) bots could be one of them.
August 18, 2020
These are stressful times. Extended times of self-isolation and sudden unemployment caused by the COVID-19 pandemic are only some of the factors provoking an increase in mental health issue cases, both in the United States and around the world.
Recent data from the Kaiser Family Foundation has revealed that over 56 percent of Americans admitted to feeling worried or stressed in relation to the outbreak.
The number grew sharply when frontline healthcare workers and their families were surveyed, with 64 percent reporting worsened mental health conditions.
The effects of these negative emotions have in turn led to at least one negative mental health effect, such as lack of appetite, insomnia, or shorter temper.
Worrying data indeed, and since nobody knows how long the pandemic will last, these numbers could very well be just the tip of the iceberg.
During these challenging times, seeking mental health help and guidance is crucial, and many therapists have swapped to etherapy to support people in need in these extraordinary times.
However, as the online therapy trend keeps growing, therapists may struggle in keeping up with patient demand, leaving some people in need of alternative forms of therapy.
Artificial intelligence (AI) bots could be one of them.
Can a bot replace a therapist? Meet Ellie
While it is difficult to give a definitive answer to this question, there are certain factors and examples that can help build a picture of the use of bots as instruments of mental health therapy.
Artificial intelligence is great at predicting patterns, and in 2017 a machine-learning algorithm was developed to help doctors find patients’ suicide attempt rates in order to prevent new ones in the future.
The AI examined huge quantities of data and connected patterns that humans might have missed—individuating which patients might have attempted suicide—with remarkable accuracy.
Crunching data aside, however, many patients would not be excited at the idea of interacting with machines that are clearly not human nor resemble a human being.
This concept brought some researchers from California to create Ellie in 2016, a virtual therapist designed to detect signs of depression and post-traumatic stress disorder (PTSD).
Ellie did not work with patients alone but acted as a data-gathering tool.
Utilizing a camera, the AI analyzed the facial expressions, head gestures, eye gaze direction, and voice quality of patients to identify behavioral indicators that could be linked with depression and post-trauma stress.
A doctor was later in charge of examining Ellie’s data and formulating appropriate treatment plans.
From data gathering to conversational bots
AI technology has evolved greatly since 2016, and bots are now able to gather even more data, more efficiently than ever.
They have also become more optimized, allowing us to run powerful bots on devices, such as smartphones and tablets.
While the role of these bots is still inherently a data-gathering one, the process is now faster, and the information collected is distilled and organized like never before.
Moreover, because of their conversational nature, some of the bots in the following list have actually been proven to reduce anxiety levels without therapist interaction.
Without further ado, then, here are the three bots that can help you deal with mental health issues.
Woebot uses cognitive behavioral therapy (CBT) to help users get better by challenging and changing unhelpful cognitive distortions and behaviors.
Created by Clinical Research Psychologist Alison Darcy in 2017, Woebot feels remarkably realistic and actually provides substantial feedback in regard to how you are feeling.
CBT is used here through a series of functions, the first of which is the Challenge Negativity one. Woebot asks users to type three different negative thoughts they are having and then encourages them to work on one.
The bot will then ask a series of questions to allow users to learn how to identify different negative thinking patterns and how to start working on challenging them.
Woebot also has a variety of additional features, including a Mood Tracker and a mode called The Power of Language, featuring videos on CBT-related skills.
Wysa has more than one million users on Android only, and its interface is clean and beautiful.
Just like Woebot, Wysa uses CBT, but the platform is more than that, offering dialectical behavior therapy (DBT), as well as meditation, breathing, yoga, motivational interviewing videos, and micro-action prompts.
The machine learning bot learns from users but also relies on the support of the Wysa team therapists.
Chatting with the bot is free, but users will have to pay $29.99 for a monthly subscription to chat with Wysa Coaches.
The user conversations that happen through the app are anonymous, but Wysa feeds them anonymized to the bot to improve its accuracy.
The app also features a variety of self-care tools, including medicine reminders, remote wellness exercises, and deep sleep training.
Last but not least, Joyable uses CBT to treat specifically anxiety and depression issues.
The bot helps users individuate a specific event triggering the negative feelings and then helps them to remove negative associations from it.
Broadly speaking, the activities within the Joyable app are divided into educational, behavioral, and thought-based.
Unlike the other bots on this list, Joyable does not offer a free option but only a seven-day trial, after which it costs $99 a month.
The price includes coach mentors who will work with users not only through the app but also over the phone, via email, and text.
The future of mental health bots
The AI bots mentioned in this list each have a track record of helping users with anxiety and other mental health issues, and their user base is growing day by day.
However, the majority of those still work through the support of human therapists, showing that bots still retain the main role of data gatherers in mental health applications.
That being said, as bot technology continues to evolve, data processing becomes more efficient and, with it, the type of help that bots are able to provide.
Will we eventually replace human therapists with AI bots and virtual assistants? Either way, it is an interesting space to watch.