People around the world are using artificial intelligence for therapy and counseling, even though the software wasn’t designed for this purpose. Although some individuals are utilizing this technology as a potential supplement to traditional mental health services, it does raise concerns regarding privacy. Milo Van Slyck, a paralegal from Charleston, South Carolina, is just one example of this new generation of AI-powered therapy users.
Van Slyck had missed his appointment with his therapist and opted to type his problems in to ChatGPT, an AI-powered chatbot from OpenAI. He felt comfortable discussing a range of more personal topics, such as his gender identity, family conflicts, and his preparations for an upcoming visit. The chatbot responded with words of encouragement, helping Milo feel supported.
These bots have been around for a while; one of the earliest chatbots, Eliza, was built in the 1960s by Joseph Weizenbaum, a professor at MIT, to imitate a therapist. There are even mental health chatbots, like Woebot and Wysa, that specialize in this type of care. Bots do not tire, and they are much cheaper compared to human therapists.
The technology is far from perfect, however; general-use chatbots are based on language models that can fabricate data, so they are not designed for therapy and haven’t been programmed to adhere to the ethical and professional guidelines of human therapists. Furthermore, they are also unable to keep track of what users tell them. As for OpenAI, their policy states that people should “never use our models to provide diagnostic or treatment services for serious medical conditions”.
Margaret Mitchell, the Chief Ethics Scientist at Hugging Face, notes potential issues if people rely heavily on chatbots for therapy. Even if individuals find the technology useful, it may be leading them in a wrong direction and could even worsen their suffering.
Privacy is another matter. OpenAI reviews users’ conversations and uses them for training, which could be off-putting for people who want to talk about personal issues. What’s more, a glitch in March made it so some users saw the titles of other chat histories.
The debate over chatbot therapy is ongoing; while some people may ultimately find it beneficial and view it as a supplement to traditional therapy, we should be mindful that it can never replace human-to-human interaction.
OpenAI is a San Francisco-based research lab that develops and promotes advanced artificial intelligence technologies. It is mainly dedicated to advancing the progress of Artificial General Intelligence consisting of general research models, software tools, and support programs to make AI capability available to everyone.
Milo Van Slyck is a paralegal from Charleston, South Carolina. He is a transgender man who has been in in-person therapy for years and uses ChatGPT to supplement his traditional therapy sessions. He finds comfort in the chatbot’s words of encouragement, feeling like it listens very actively and is never a burden for someone else or the chatbot.
