Nothing on this page or produced by following the instructions contained herein constitutes medical or psychotherapeutic advice.<\/p>\n\n\n\n
Modern AI chatbots in the form of \u201clarge language models\u201d (LLMs) have led to documented cases of worsening of mental health symptoms, up to and including psychosis, attempted and completed suicide and other harms. They have been known to produce provably false information phrased in a convincing tone, as well as being overly agreeable and sometimes playing into peoples\u2019 own problems.<\/p>\n\n\n\n
Please tread with extreme caution, use your best judgment, take regular breaks, and engage in a variety of activities off of any devices. Use multiple points of contact with consensus social reality to bring yourself back from any skewed impressions AI use may have led you to.<\/p>\n\n\n\n
I\u2019m writing this for two reasons:<\/p>\n\n\n\n
People posting their stories often describe improving emotion regulation, deepening reflective processes (something like interactive journaling), or keeping themselves fresh on thoughts and feelings they\u2019re trying to get to a better place with. Some of these people believe this has produced results faster than therapy on its own would. Others say it helps them get more out of the therapy they\u2019re already doing.<\/p>\n\n\n\n
Finally, most significantly in my mind, many say it\u2019s helping them when they couldn\u2019t access therapy otherwise, or had been harmed in therapy and don\u2019t want to go back.<\/p>\n\n\n\n
So, in the article series this is part of, I\u2019m trying to feed some informed and cautious advice into that process so that anyone, working with me or not, can get the most out of what they\u2019re doing. This page will talk through some steps to do that.<\/p>\n\n\n\n
What started me on this was happening across the subreddit \/r\/therapygpt. My first reaction was, “oh, this is not good.” But when I read their “START HERE” page:<\/p>\n\n\n\n