Chatbots and the Future of Psychoanalysis

Chatbots and the Future of Psychoanalysis 

Jan Jaramillo

Psychoanalysis as a psychotherapy technique is dying; it is either not accessible, expensive, or sometimes too long to complete the treatment. In today’s chaotic society, it is difficult to even find time to introspect in one’s free time. Without much-needed access to psychotherapy, treatments for even common pathologies like depression or anxiety are becoming increasingly difficult as patients have less resources and time to devote to their mental health. A study done by Dr. Taylor Warren and Dr. Olusa Ajilore in 2024 demonstrated how, even with antidepressant treatment and psychotherapy, there is still a lack of biomarkers to be able to predict or prevent relapse in patients suffering from late-life depression (LLD). Dr. Ajilore then presented a study he is a part of that uses chatbots instead of face-to-face psychotherapy to target the uncinate network; this network consists of regions in the frontal lobe connected by a white matter tract, involved in many disorders. The disruptions of white matter in this network are associated with depression. Reduction in network efficiency predicted response to cognitive behavioral therapy (CBT); as connectivity got stronger, connectivity between regions in this network led to better responses to CBT. Many networks displayed subsets of patients who had better responses to different therapies based on which network was affected. People who did eight sessions with the chatbot they made (LUMEN) showed greater efficiency in the targeted network. Psychoanalysis offers an alternative or promoter to CBT through the usage of free association. Here, the patient is left to speak freely with no interruption about miscellaneous topics until a clear pause is reached in which the psychoanalyst reflects on what has been said; connections here are drawn between topics that associate without conscious thinking. For example, a patient could be speaking about butterflies and how they appreciate their flight and freedom to fly from flower to flower. The patient then shifts to a more personal topic about how they feel frustrated with how controlling their partner is; a clear association is drawn between the freedom they seek and how their relationship restrains them from achieving it. The psychoanalyst then shows the need to explore the unconscious reasons or conflicts of the patient, which are limiting their ability to be independent. The goal is to suffer less and enjoy more through self-knowledge and acceptance. The key to keeping psychoanalysis alive may lie in this easy-to-access chatbot. A study has been done by Leuzinger-Bohleber, Marianne et al. (2019) that showed no significant difference between psychoanalytic therapy and CBT: meaning that patients may be able to respond the same, if not better, to psychoanalysis depending on their neural and psychological frame. A significant issue may arise in the fact that it is not face-to-face therapy; however, Psychoanalysis relies on concepts like transference, which requires a strong, regular, often strong frame between the psychoanalyst and the patient. Through Dr. Ajilore’s study, however, there may be hope to simulate this human connection. The study conducted using LUMEN showed that non-white women with college degrees or less had the greatest responses to the treatment; the people who helped design the chatbot were non-white women. Somehow, their human traits managed to significantly transfer through LUMEN to show a significant difference in connection, and therefore in treatment, with those that shared the same background. Similarly, one could theorize that an algorithm could be developed to simulate this same connection, or adapt to create it, depending on the patient. Just as a patient may not trust their caregiver, the robot could not only find a way to create that trust, but maybe highlight the patient’s inability to create trust in their own life. Psychoanalysis is also long, but with easier access, less expense, and more sessions, more work can be done in a shorter time to achieve a suitable treatment plan. This can be additionally boosted by adding functions like scanning and measuring for voice variation, body language, topic subjects: creating a framework for the patient’s defense mechanisms. These markers could be the missing step in predicting and preventing relapse, not just for LLD but other psychopathologies. 

https://pmc.ncbi.nlm.nih.gov/articles/PMC6364135/

Taylor, Warren D, et al. “Reconsidering Remission in Recurrent Late-Life Depression: Clinical Presentation and Phenotypic Predictors of Relapse Following Successful Antidepressant Treatment.” Psychological Medicine, 8 Jan. 2025, pp. 1–12, https://doi.org/10.1017/s0033291724003246. Accessed 16 Jan. 2025.