There is nothing but a knot of discomfort in me every time I think about relying on an AI-based recommendation rather than my own clinical intuition – and the very discomfort was the result of Reamer’s talk about the problem of social workers’ dependence on AI. As a person who likes technologies and sees their ability to make processes more efficient, I know too well how simple it might be to allow myself to let a nicely structured answer provided by the system become the ultimate conclusion without giving it any thought.

I haven’t used AI in my clinical practice yet while interning at Highland Rivers Behavioral Health. Still, I use it all the time in school to organize my thoughts, make sure the text looks good, and find the necessary data. From those experiences, I learned two lessons: AI really saves one sometimes and is usually confidently wrong or unaware of the context more often than not. And in the world of social work, awareness of the context is crucial – culture, trauma background, strengths of the client, and many other aspects of his or her life can’t be captured within a pattern in a dataset.

I concur with Reamer on the point that artificial intelligence is never to replace professional judgments, informed consent, or the making of ethical decisions; clients have a right to know when and how the use of technology influences their treatment. The only point that makes me disagree a bit with Reamer in his statement is tone; I believe that the use of artificial intelligence is never to be viewed from a negative angle.

If I were to explain this question to my client today, I would tell them that while I may rely on technology to assist me in organizing information or working more efficiently, I would never allow the computer to make decisions on what is best for you. The point of therapy is to hear your story, learn about your situation, and to make decisions ethically according to professional standards, and not the computerized one.

Part 2

While Haidt and Rausch make their arguments based on teenagers, I have observed how well their evidence holds even for adults during my internship at Highland Rivers Behavioral Health. The seven pieces of evidence provided by the authors helped me frame things I have seen among the clients in terms of a broader understanding.

There are certain groups that I feel are being greatly influenced by technology. One such group of individuals would be adults who are under treatment for addiction and mental illness. In one such case, I have found clients admitting to spending time on TikTok and Instagram late at night to divert themselves from cravings or isolation, ending up feeling even more anxious when comparing their lives with pictures on those sites. Moreover, individuals in early stages of recovery find that their work on cutting off their using contacts gets undone through sites such as Facebook. Principles of techno skepticism proposed by Haidt seem to imply that screen usage should be considered as important as sleep or substance use while carrying out an assessment.

It applies to couples and families as well. In couples therapy, I have witnessed tensions between partners over Instagram likes or messages, with feelings of jealousy and comparison arising, which makes sense from the standpoint of the ultra-social model described by Haidt, because those websites prey on precisely the wiring that he talks about. With regards to families, I have met many parents who are frustrated by how their children interact with their phones, while acknowledging they themselves check their phones during family dinners.

It becomes even more difficult when it comes to working with isolated individuals – such as my clients suffering from disabilities or various illnesses who form connections through online networks such as Facebook. To those individuals, the same website that Haidt is criticizing becomes a means of establishing connections rather than a source of danger. Therefore, I do not believe the aim is to dissuade people from using technology; it should be about helping them analyze what works for them specifically.