Like all technology, artificial intelligence has the potential to solve great problems … and also the ability to cause new ones. To the current controversy about the use of AI now have to add a new one that, for now, It does not have a clear positive or negative ending: its impact on mental health.
According to a study published in Mental Health and led by Dorian Hatch, when the responses written by psychotherapists are compared with those of Chatgpt, the latter usually have a higher score.
The question of whether the machines can be therapists is an issue that has received greater attention due to some of the benefits of working with generative artificial intelligence (AI). It is increasingly obvious (and disturbing) to discover that an AI can write with empathy and that the generated content receives a high rating both by mental health professionals and volunteer service users, to the point that it often prefers it to the content written by professionals.
In the new study, which more than 800 volunteers participated, the hatch team showed that, Although differences in language patterns were noted, people could rarely identify whether the answers had been written by Chatgpt or therapists.
This finding is one more test of Alan Turing’s prediction: The time will come when we will not be able to distinguish between the answers written by a machine and those that belong to other humans. In addition, the responses written by ChatgPT were generally qualified higher in the basic rector principles of psychotherapy.
According to the results, the responses generated by ChatGPT were generally longer than those written by therapists. After controlling the length, Chatgpt continued to respond with more nouns and adjectives than therapists. Taking into account that nouns can be used to describe people, places and things, and adjectives can be used to provide more context, this could mean that Chatgpt contextualizes more widely than therapists. A more extensive contextualization may have led respondents to qualify the highest chatpt responses in the common therapy factors.
According to the authors, these results can be An early indication that Chatgpt has the potential to improve psychotherapeutic processes. In particular, this work can lead to the development of different test methods and creation of psychotherapeutic interventions.
Given the growing evidence that suggests that generative AI can be useful in therapeutic environments and the probability that it can be integrated into therapeutic environments before thinking, The authors ask mental health experts to expand their technical literacy to ensure that the AI models are trained and supervised carefully by responsible professionals, improving asYo quality and access to attentioneithern.
“Since Eliza’s invention almost sixty years ago, scientists have debated if AI could play the role of a therapist -concludes the study -. Although there are still many pending important questions, our findings indicate that the answer can be “yes.” We hope our work drives both the public and mental health professionals to ask important questions about ethicsthe viability and usefulness of integrating AI and the treatment of mental health, before the AI train leaves the station ”.