AI Readings for Clinical Psychologists and Psychiatrists

While writing about the history of the Cognition and Affect project, I received a request for some readings on AI and psychotherapy. So, I thought I’d share a few readings here. This document is not comprehensive; it is biased towards our projects.

One of my main professional goals is to help people use knowledge resources and information technology, in a manner that is informed by broad (affective) cognitive science, to develop themselves. The ensuing projects are relevant to much adult deliberate learning, self-help, and psychotherapy (as noted below.)

An example of this was the very extensive Learning Kit project at SFU, whose principal investigator was Dr. Winne, and for which I was the technical lead. We developed an app and a web service to help learners: gStudy and nStudy. That software is also an educational psychology research tool: to study self-regulated learning. (Phil Winne is the Canada Research Chair in Self-Regulated Learning & Learning Technologies.) The software we develop at CogSci Apps Corp. is inspired by this approach. I.e., our apps (a) are based on cognitive science, (b) are designed to help end users, (c) are also research tools. The research results from our apps feed back into the design of our products (and into our understanding of course). This is a modern, cognitive science-based analog of the “research practitioner” model in psychology.

Some Readings

Anyone interested in AI applications for psychotherapy should read:

  1. The Computer Revolution in Philosophy (Sloman, 1978),
  2. “You don’t need a soft skin to have a warm heart: Towards a computational analysis of motives and emotions” Sloman & Croucher (1981), and
  3. “How many separately evolved emotional beasties live within us?” Sloman (2003).

These documents don’t delve directly into clinical applications; but they contain essential foundational ideas.

My Ph.D. thesis was on motivation from an information processing perspective. It extended and implemented ideas developed by Aaron Sloman on the Cognition and Affect project. That theory provides a novel account of (tertiary) emotions as “perturbances”. Perturbance is a very important, but quite subtle, architecture-based concept. That means that the concept can only be understood with respect to an architecture of mind. (Aaron Sloman was the first researcher to systematically explore mental architectures in AI.) Chapter 6 of my thesis suggests that obsessive-compulsive disorder can be understood in terms of our theory.

The same year I published my thesis, Adrian Wells and Gerald Matthews published Attention and emotion: A clinical perspective. Their book is similar in spirit to my approach. For example, they referred to AI architectures; they emphasized the relations between attention and emotion; and they pointed to the relevance of these concepts to clinical psychology. Ironically, they were in England at the same time I did my research there, but our paths didn’t cross and we didn’t reference each other then (the days before the web.)

Marvin Minksy’s The Emotion Machine contains many big ideas for thinking about the mind as a collection of mechanisms. (Minsky passed away last week. Hopefully, the hommages will lead to a resurgence in the kind of AI that he conducted: focusing on big questions, big ideas and whole minds. Here’s a link to him talking about Cognitive Architectures and Aaron Sloman’s work.)

In 1996, Ian Wright, Aaron Sloman and I published “Towards a design-based analysis of emotional episodes” in Philosophy, Psychiatry, & Psychology. This paper applies our AI concepts to grief.

My first book, Cognitive Productivity: Using Knowledge to Become Profoundly Effective describes the scientific basis and several practical applications of my approach. I argue that deep learning needs to involve many different aspects of the “architecture” of the human mind. Part 2 describes an AI software architecture that is relevant to this. That is a variant of Aaron Sloman’s “CogAff” architecture. Chapter 15 briefly describes applications of my approach to psychotherapy and self-help. In my view, psychotherapy is a form of education; and education is a form of psychotherapy.

Insomnia is, of course, quite important to psychotherapists and psychiatrists. Insomnia is often a consequence of psychological challenges. It can also lead to psychological concerns. So, I have designed (and we are continuing to assess and improve) new cognitive treatments for insomnia, including serial diverse imagining. mySleepButton® is an app that implements this treatment. You’ll notice references to AI literature in my 2013 paper, “The possibility of super-somnolent mentation: A new information-processing approach to sleep-onset acceleration and insomnia exemplified by serial diverse imagining”. We continue to develop this app. Nancy Digdon, colleagues and I will soon publish empirical results on our treatment that used SomnoTest, a research version of the mySleepButton app.

CogSci Apps Corp. will soon release an update to mySleepButton that includes techniques from acceptance and commitment therapy. We will then roll-out additional cognitive science-based features.

It should be noted that mySleepButton is also an emotion regulation app. It may help some users regulate their current perturbance(s), i.e., to decrease the insistence of intrusive motivators.

While I’m on the subject of perturbance, I would highly recommend to anyone interested in clinical applications of AI models of affect to get the hang of this concept. It is increasingly recognized that repetitive thought, rumination, and obsession are, as Wells & Matthews suggested, key indicators of psychological functioning. (I recognize that’s a vague statement; but I’m not writing a research paper here. See Watkins, 2008, on “Constructive and unconstructive repetitive thought”.)

I also have published a brief AI-inspired document about my project to apply acceptance and commitment therapy, cognitive science and technology to develop an emotion regulation app: “Specification for a productive practice app to assess and improve psychological treatments for romantic grief and other tertiary emotions”. (There are obvious connections with mortal grief, and hence the Wright, Sloman and Beaudoin paper mentioned above.)

Dr. Dean Petters and I are currently co-authoring a chapter on computational modeling of attachment for a book on Computational Psychiatry. Dean Petters is one of a rare breed: mental health researchers who extend and apply AI literature. Check out of his publications here http://www.cs.bham.ac.uk/~ddp/. He teaches mental health courses in universities and also gives private seminars on the subject.

As I alluded to elsewhere on this site, quantitative modelling (connectionism, Bayesian models, dynamic systems, …) is all the rage in AI — continuing a trend that started in the 1980s. I believe the other chapters in the upcoming Computational Psychiatry book will be in that trend. Currently in AI, there is not much attention given to agent level architectures that attempt to do justice to the major categories of affect (motive processing, emotions, moods and attitudes). (See however the Biologically Inspired Cognitive Architectures society.)

Dr. Eva Hudlicka also models emotional phenomena, and she also applies and extends AI to psychotherapy research and to her own clinical practice (http://therapy21st.net). She is a lecturer at the College of Information and Computer Sciences of the University of Massachusetts. At CogSci-2014, she chaired a workshop on “Computational Modeling of Cognition-Emotion Interactions: Relevance to Mechanisms of Affective Disorders and Therapeutic Action”.

Boden’s (2006) voluminous Mind as Machine: A History of Cognitive Science. Its index will be your friend, with several indices for psychopathology, psychotherapy and emotions.

Dr. Rick Cooper’s work on habits and cognitive control (in terms of contention scheduling) is also pertinent to clinical applications. Most of what Cooper and Tim Shallice have written about clinically concerns neuropsychology. Like many members of the Cognition & Affect project, these researchers believe that a computational understanding of mind can shed light on clinical concerns.

Merlin Donald’s A Mind So Rare does not describe clinical applications. Nor does it explicitly delve into AI. However, it presents an architectural approach to the human mind that has much in common with the CogAff architecture mentioned above.

References


Beaudoin, L. P. (2015, July). Specification for a productive practice app to assess and improve psychological treatments for romantic grief and other tertiary emotions. Poster presented at ISRE 2015. Geneva, Switzerland. Retrieved from http://summit.sfu.ca/item/15224.

Beaudoin, L. P. (2015), Cognitive Productivity: Using Knowledge to Become Profoundly Effective. BC: CogZest.

Beaudoin, L. P. (2013). The possibility of super-somnolent mentation: A new information-processing approach to sleep-onset acceleration and insomnia exemplified by serial diverse imagining. Cognitive Productivity Research Project, Simon Fraser University. http://summit.sfu.ca/item/12143

Beaudoin, L. P. (2014, July). A design-based approach to sleep-onset and insomnia: super-somnolent mentation, the cognitive shuffle and serial diverse imagining. Paper presented at the 36th Annual Conference of the Cognitive Science Society, workshop on “Computational Modeling of Cognition-Emotion Interactions: Relevance to Mechanisms of Affective Disorders and Therapeutic Action”, Québec, Canada.

2014 Annual International Conference on. Biologically Inspired Cognitive Architectures (BICA). Biologically Inspired Cognitive Architectures

Beaudoin, L. P. (1994). Goal processing in autonomous agents. (Doctoral dissertation). University of Birmingham, Birmingham UK. Retrieved from http://www.sfu.ca/~lpb/tr/Luc.Beaudoin_thesis.pdf

Biologically Inspired Cognitive Architectures Society.

Boden, M. A. (2014). Aaron Sloman: A bright tile in AI’s mosaic. In From animals to robots and back: Reflections on hard problems in the study of cognition (pp. 9-30). Springer International Publishing.

Boden, M. A. (2006). Mind as machine: A history of cognitive science (2 volumes).

Cooper, R. P. (2010). Cognitive control: Componential or emergent? Topics in Cognitive Science, 2(4), 598–613. http://doi.org/10.1111/j.1756-8765.2010.01110.x

Cooper, R. P., & Shallice, T. (2006). Hierarchical schemas and goals in the control of sequential behavior. Psychological Review, 113(4), 887–916; discussion 917–31. http://doi.org/10.1037/0033-295X.113.4.887

Cooper, R. P., & Shallice, T. (2000). Contention scheduling and the control of control of routine activities. Cognitive Neuropsychology, 17(4), 297–338. http://doi.org/10.1080/026432900380427

The Cognition And Affect Project web site at the University of Birmingham. http://www.cs.bham.ac.uk/research/projects/cogaff/

Dawkins, R. (1976). The selfish gene. Oxford, England: Oxford University Press.

Donald, M. (2001). A mind so rare: The evolution of human consciousness. New York, NY: W. W. Norton & Company.

Hudlicka, E. (2014). Affective BICA: Challenges and open questions. Biologically Inspired Cognitive Architectures, 7, 98–125. http://doi.org/10.1016/j.bica.2013.11.002

Hudlicka, E. (2008, March). What are we modeling when we model emotion?. In AAAI Spring Symposium: Emotion, Personality, and Social Behavior (pp. 52-59).

Minsky, M. L. (2006). The emotion machine: Commonsense thinking, artificial intelligence, and the future of the human mind. Hew York, NY: Simon and Schuster.

Petters, D. & Beaudoin, L. P. Attachment Modelling. Manuscript in preparation; to appear in Computational Psychiatry.

Sloman, A. (2003). How many separately evolved emotional beasties live within us? In R. Trappl, P. Petta, & S. Payr (Eds.), Emotions in humans and artifacts (pp. 35–114). Cambridge, MA: MIT Press.

Sloman, A. (1978). The computer revolution in philosophy: Philosophy, science and models of mind. New York, NY: Harvester Press. Retrieved from
http://www.cs.bham.ac.uk

Sloman, A., & Croucher, M. (1981a). Why robots will have emotions. Proceedings of the 7th Int. Joint Conf. on AI.

Sloman, A., & Croucher, M. (1981b). You don’t need a soft skin to have a warm heart: Towards a computational analysis of motives and emotions (No. 004). Cognitive Science Research Paper, Sussex University. http://www.cs.bham.ac.uk/research/projects/cogaff/81-95.html#55

Watkins, E. R. (2008). Constructive and unconstructive repetitive thought. Psychological Bulletin, 134(2), 163–206. http://doi.org/10.1037/0033-2909.134.2.163

Wells, A., & Mathews, G. (1994). Attention and emotion: A clinical perspective. Hillsdale, NJ: Lawrence Erlbaum Associates Publishers.

Wright, I., Sloman, A., & Beaudoin, L. P. (1996). Towards a design-based analysis of emotional episodes. Philosophy, Psychiatry, & Psychology, 3(2), 101–126. doi:10.1353/ppp.1996.0022

Published by

Luc P. Beaudoin

Head of CogZest.
Author of Cognitive Productivity .
Cognitive productivity consultant and public speaker.
Adjunct Professor of Education & Adjunct Professor of Cognitive Science, Simon Fraser University
Co-founder of CogSci Apps Corp.
See About Me for more information.

Leave a Reply

Your email address will not be published. Required fields are marked *