Читать книгу The Future of Health - Roberto Ascione - Страница 25

Monique Levy

Credit: Monique Levy – Private archive

Relational agents represent the most profound shift in digital health to date, and the most exciting application of AI, conversational UX, and other technologies combined.

Relational agents are software designed to build and maintain a social-emotional bond with users over time, and to enhance or help the user in some way, such as shifting cognition and affect to new behaviors, decisions, and knowledge. The concept of a relational agent was first discussed more than 20 years ago, most notably at MIT Media Lab's Affective Computing Group, which developed and evaluated an exercise advisor system to explore interactions between people and relational agents. The lab's study¹ found user ratings along the lines of respect, trust, and likeability were higher for the relational agent than those for the non-relational agent. A decade later another study² demonstrated that “individuals were more willing to disclose to an artificially intelligent ‘virtual therapist’ than when they believed it was human-operated.”

Today's relational agents are capable of so much, but one of their more exciting applications is in the area of mental health, and one of the most innovative companies in the space is Woebot Health, where I am the chief commercial and strategy officer.

Woebot Health is the company pioneering industrializing relational agents for various unmet needs in mental health, and the first to apply them to CBT-based therapy. Its relational agent, Woebot, is accessible via apps that use a conversational UX. Woebot is built from core principles of human relationships and therapeutic process and enables people to more deeply engage with their mental health in the moment, and stay engaged as new episodes emerge over a lifetime. Woebot is also capable of quickly building a bond with users. Bond, or therapeutic alliance in clinical terms, reflects the rapport that develops between participants and Woebot. In psychotherapy, close rapport between therapist and client is a predictor of the level of change achieved. Woebot has recently demonstrated an ability to show human-level bond in a large-scale study using a standard validated measure of therapeutic relationship. Results show that Woebot can quickly create a bond non-inferior to human therapists, and maintains that bond over time, which can lead to improved mood, deeper engagement in treatment, and improved outcomes.

Woebot was created by Alison Darcy, PhD, a clinical research psychologist and health-tech visionary who believes in technology's ability to unlock human potential. Dr. Darcy built her first digital mental health solution way back in 2001. Since then, she has built an international reputation for her scientific research. At Stanford, Dr. Darcy worked with AI pioneer Andrew Ng to explore the intersection of AI and health-care, leading to his Health Innovation Lab in Computer Science. And as founder of Woebot Health, she formed the foundational ideas for an on-demand digital therapist that can support people across the lifespan. Now, Woebot is at the heart of Woebot Health's next-generation digital therapeutics for anxiety and depression, which provide the scalable innovation that can deliver trusted and effective mental health care to millions of people simultaneously.