In this final newsletter of the year, we are introducing a specific section for discussion of issues that need clarification of their place in the Mental Health field.
In Newsletter 89 a note on “Psychotherapy” had comments on several ways that modern technology is being introduced to assist in case management. Whilst no comments were elicited by description of automated methods of providing information, issuing reminders and recording of messages and results of assignments, the one matter that did attract comment was on the use of Artificial Intelligence software to provide empathic responses to clients contacting the message bank with issues of concern arising between scheduled appointments.
The gist of the comment was that empathy is a human characteristic. The centrality of empathic responses in the therapist-client relationship raised doubts that AI could emulate this. Whether or not it could is a reasonable question. Whether or not it needs to is also a reasonable question. Being a human characteristic does not preclude it being emulated.
There are many human characteristics, such as judgment of probabilities and likelihood of consequences, that are quite adequately provided by AI. The success of AI programs for reading X-Rays to detect breast cancer is a case in point. Rather than excluding AI by definition, it is incumbent on us to examine programs to see if they work. Allan Mawdsley Our President, Jo Grimwade, points out that the use of Ai is an application of algorithms.
An algorithm is “a process or set of rules to be followed in problem-solving operations, especially by a computer”. He comments “In defence of algorithms and the problematic application to psychotherapy”: “Maybe we did not know that we use algorithms frequently and repeatedly, but we do. Planning a busy day, following a recipe, and writing a paper all involve the invoking of well practiced algorithms. So does psychotherapy.
Etymologically, algorithm is an Arabic word derived from the name of a ninth century scholar, Muhammad ibn Musa al-Khwārizmī, meaning, “native of Khwarazm” (modern Khiva in Uzbekistan). I think it is possible to try to pronounce his name and end up with an Anglicized version that sounds like algorithm. He wrote a book that translates as “the compendium on calculation by restoring and balancing”, it included a section called al Jabr, “the reunion of broken parts”, which is now called algebra.
There are three important distinctions to be made in contemporary usage: deterministic algorithms, inferential algorithms, and heuristics. The last concerns ways of understanding and much of psychotherapy would involve this non-conclusive style of thinking that generates potential insights: “let me understand what you just said …?” Inferential algorithms are what drives the endless list of unwanted emails that come when one expresses an interest in a product or event: “because you liked this, you may like these other sorts of things?” Again, this would be a mode of thinking in psychotherapy: “your hatred of your female boss may be connected to dissatisfactions with your mothering”. An inference that might open up insight.
Deterministic algorithms are also part of psychotherapy, but more in the form of process management: “what brought you here today? On what have you reflected since our last session?” The intention is to establish clear facts and details.
Algorithms drive so-called artificial intelligence. So many comments by a psychotherapist are repeats of things said to others. This could be thought of as robotic, but there is a lot of efficiency in well-tried phrases. After all, one needs only to be close to the relevant point to enable the patient find the truly insightful part of what has transpired. In fact, being accurate can erode the effectiveness of the psychotherapist over time as the patient comes to depend on the psychotherapist to provide answers.
I find supervision to be even more algorithmic. I will suggest what might be said to the learning practitioner. Often they will write down slavishly my exact phraseology. Eventually, they will see that my wording is more heuristic than deterministic, but not initially.
Dream analysis can have a very algorithmic process. “Please repeat your dream … just the dream without elaboration … who is in the dream? Do you recognize anybody or anything … how does the dream end? … now what do you make of it?” However, the differences between psychotherapy as applied algorithms and psychotherapy proper are important to observe.
In psychotherapy, many themes appear at once: choosing the most pertinent theme and then being able to return to other themes is a skill of parallel thinking. Some of the themes will be of algorithmic form, others will be unexplored oddities that can only become algorithms after some processing.
Secondly, the algorithm is subject to the influence of the observation of self or other. This is Heisenberg’s Uncertainty Principle (HUP). The algorithm can be masked by interaction of observing with observed. In psychotherapy, this happens all the time. Thirdly, human to human exchange is not reducible to algorithmic thinking. Care is part of something that establishes communication at a level of respectfulness. Listening is more than hearing.
Finally, the question of the transference! Standard transferences about mother and father may be rendered by algorithm, but the subtlety of the affect cannot be captured by one toned transference interpretation. There is always multiple shades of meaning in transference, including: “will you believe me if I say what I need to say? …”
On the other hand, AI may make it possible for someone to start, impersonally, on a psychotherapy journey. I believe that some AI counselling programs do serve this function of referring on to an experienced practitioner. We use algorithms all the time. We need to be wise to our usage.”
Jo Grimwade,
Reader’s comments would be welcome.
From the report in our last newsletter of our Strategic Planning it can be seen that we need help with a variety of tasks.
Some readers may already have an interest in these tasks and would be willing to help us without adding to their workload (because they are already doing the task) and without needing to belong to our committee or doing much more than they are already doing.
For example, we know that there are people who keep watch on the websites of the departments of community services and health for announcements of programs and tenders for projects. MHYFVic really needs to keep up to date with these, but we do not have anybody to do the watching. A volunteer reader who could alert us to such news would be hugely helpful.
Similarly, we would love to have someone alerting us to current preoccupations of young people on social media. Our committee is just not in touch with those things, but it would be great if someone could bring us up to date.
At a deeper level, we also need help with legal advice and public relations expertise.
Allan Mawdsley, Vice-President, MHYF Vic
The October Newsletter has a link to a Webinar on “Working therapeutically with children who have experienced trauma from physical or sexual abuse”.
There is also a link to a podcast on “Responding to shame and stigma in child protection and reunification”.
The MHYFVic Workshop we were planning to hold on 26th October had to be unexpectedly cancelled. We hope to hold it next year, all being well. The reason was that the venue, which is a facility of the University of Melbourne, was suddenly informed that new security measures were to be put in place for public events, but the rules had not yet been established. It appears to have been a casualty of the recent anti-Israeli protests. We will keep readers informed of developments.
The MHYFVic Bulletin we sent out to our liaison network collaborators this month focused on the dysfunctional child protection system. A year ago, we requested a meeting with the Minister to propose a pilot project to improve client outcomes and worker satisfaction but received a brush-off.
In the year gone by a further 66 children in the care of the Department have been killed. We have again asked to meet with the Minister.
MHYFVic members propose to present the case for a national body representing the interests of young people and their families in the mental health advocacy field because we have consistently been sidelined by the adult “noisy wheels”.
Mental Health for the Young & their Families in Victoria is a collaborative partnership between mental health & other health professionals, service users & the general public.
MHYFVic
PO Box 206,
Parkville, Vic 3052