Active Inference-based Psychotherapy: What might psychotherapy based on first principles look like?
Working notes on a potential approach to psychotherapy and growth.
Copyright Grant H Brenner
NB This is not medical advice nor an established approach to psychotherapy. If you are concerned about your well-being or are in need of assistance, please seek professional evaluation right away. For crises, the 988 Suicide & Crisis Lifeline provides immediate support and resources.
Summary & Overview
Phase 1 — Capture
- Conducts initial assessment using conventional methods plus AIT-specific evaluations
- Evaluates patient’s self-modeling capabilities and personality structure
- Determines appropriate communication style and technical vocabulary level
- Establishes hierarchical therapy goals
- Begins forming therapeutic relationship (mathematically/engineering “coupling”); Interactive Inference begins
- Assesses patient’s ability to monitor internal states and update mental models
Phase 2 — Initial Evidencing
- Implements 1–3 initially selected approaches
- Alternates between structured inquiry and open exploration
- Tracks responses using Active Inference measurements
- Refines initial diagnosis for better precision
- Handles higher model uncertainty as patient experiments with new approaches
- Updates approach based on early outcomes; Interactive Inference deepens
Phase 3 — Working Phase
- Consolidates understanding gained from earlier phases
- Develops specific action plans and goals
- Makes ongoing adjustments based on progress tracking
- Increases patient’s mastery of new models and behaviors
- Monitors for need to return to earlier phases during major life changes
- Builds greater certainty in patient’s updated mental models
Phase 4 — Release
- Plans optimal termination process (gradual taper vs. set date)
- Reviews and consolidates therapeutic gains
- Strengthens patient’s self-monitoring abilities
- Establishes post-therapy monitoring parameters
- Ensures patient has internalized model-updating skills
- Prepares for transition out of therapeutic relationship
Phase 5 — Post-termination
- Implements structured monitoring if appropriate
- Tracks potential recurrence of issues
- Maintains plan for handling unpredictable challenges
- Allows for return to therapy if needed
- Continues patient’s independent practice of skills
- Evaluates long-term effectiveness of model updates
"Active Inference Therapy" - What Might a Psychotherapy Based on First Principles Look Like?
Description: Active Inference presents a powerful framework for applying Bayesian epistemology to a variety of fields. Going beyond complexity theory, building upon it with computational models, opens up possibilities.
As “self-evidencing creatures” (Karl Friston — see below for podcast interview with Dr. Friston), we move through indeterminate spaces seeking greater adaptation. We have a model, and we use it. That model tends to be hidden, or implicit. Quite often the conscious reasons we think we make decisions are not actually the real reasons.
It shapes our perceptions and our actions. We have a prior belief about that model and then when we get new evidence as a result of our actions or the way our perceptions lead us to make choices, we then want to update that model. That is the posterior. In Bayesian logic, the prior and the posterior probabilities are related via Bayes Theorem. This basic type of logic, conditional probability, has been presented as the basis for mind, brain, and behavior.
Let’s say you guess that it will rain today with a 70% chance. Later on, you learn that the sky is dark and it feels humid. You update your guess to a 90% chance. Your prior (70%) has been updated to a posterior (90%). The probability of A (rain), given B (now I know what the sky and air are like) is given by Bayes Theorum.
Let’s say I know I generally trust people. On average, I’ll trust people 90% of the time — my prior. But then I learn someone has lied to a friend in the past about something relatively minor or ambiguous (my certain is low — there is a high prediction error). I then change my guess about them to a 60% trust (posterior probability)— and presumably continue to gather evidence to keep my model up to date.
Therapy-Exploring a Co-Constructed Higher Dimensional Space
In therapy, the space explored conceptually is at least partially liminal - between interoceoptive and exteroceptive. As co-self-evidencing creatures, the patient-therapist dyad explores this "psychoceptive" or "psycheceptive" space, using data external to the therapy, as well as internal to the therapy, to update priors and refine how updating takes place.
The dyad itself becomes an agentic system, putatively with its own ontological status - the "analytic unity" - drawing upon autopoiesis terminology and with recognition that the "frame" of therapy is a crucial boundary condition, putatively a Markov Blanket. Mathematically, such a boundary contains within itself all the data needed to understand causality.
Given this - and granted that all forms of psychotherapy involve two individuals exploring an uncertain space together to accomplish any one of a number of goals, from relief of specific symptoms or accomplishing pre-determined goals, to a more general developmental process - Active Inference has the potential to add specificity to existing models, as has been and is being elaborated.
Any specific instance of therapy also has to be grounded in the unique individual and dyadic factors. Some of these are shared generally, such as how the brain operates on its basic schematic, which also may be broad classes e.g. neurotypical versus neurodivergent — while maintaining some consistency for all people, including some aspects of affective neuroscience (how emotions work) as well as information processing in the brain, cognitive functions, etc. Neuropsychoanalysis tells us that individuals will have unique characteristics, however, which need to be taken into account. Likewise, different clinical conditions are associated with significantly conserved neurobiological findings, and that needs to be taken into account in more specific cases, or even subtypes of therapy. We are not suggesting a “one size fits all” approach, but rather a metaframework.
Beyond providing explanatory power to existing models, however, this exploratory presentation will take a high-level approach to conceptualizing what a model of therapy - Active Inference Therapy or Active Inference-Based Therapy - drawing upon active inference and the free energy principles - might look like based on those first principles.
— Sigmund Freud, 1937 —
"In another group of cases we are surprised by an attitude in our patients which can only be put down to a depletion of the plasticity, the capacity for change and further development, which we should ordinarily expect.” (Analysis Terminable and Interminable)
“One cannot perceive what one cannot classify.” — Eric Kandel
“Operationalism is based on the intuition that we do not know the meaning of a concept unless we have a method of measurement for it. It is commonly considered a theory of meaning which states that ‘we mean by any concept nothing more than a set of operations; the concept is synonymous with the corresponding set of operations’ (Bridgman 1927, 5).” https://plato.stanford.edu/entries/operationalism/
An active inference free energy based model of psychotherapy necessarily cannot favor any one particular prior form of psychotherapy, or particular practice adjacent to therapy, such as meditation or body work.
Rather by definition the active inferential process must be able to be as free as possible to apply wherever it is needed to update performance. Performance may not be the best word to use here but it is being used to refer both to the formally defined improvements in efficiency defined underactive inference and as reflected in real world outcomes as well as subjective experience, and the harder to define experience of navigating one’s own internal landscape including as that reflects the social environment. Fundamentally active inference is based on the ability to move most effectively on a broad landscape in a parsimonious way for the well-being of the individual or group depending on how the agentic envelope is drawn.
On a higher order level, the active inference process is going to be improved over the course of psychotherapy and will continue to maintain and improve Itself by definition after the period of intervention stop. As with other forms of therapy it will include refining self monitoring in order to identify the need for additional work.
Process Model
Developing AGI tools to support process and therapist could be useful for every phase.
Improve perceptual tools
Dyadic calibration
Improve process
Refine actions in psychic space — map out order of operations as hypothesis and define multiple models for different levels of modeling — map out simulated personalized developmental sequences (SPDS) and test, update, re-test (may be more or less conscious) (insert illustration)
Translate to reality
Therapist uses meta-awareness (e.g. countertransference in Psa) to detect change in uncertainty in patient’s affective arousal and by extention autonomic arousal in order to optimize precision inference and support model updating.
Contemplative practice, body work and learning to trust body and relation to arousal coherence. In trauma (sexual abuse, bullying, infirmity), one often feels betrayed by one’s body.
Identify tools e.g. affective arousal overall level and levels of naming of emotions, requires convention within each therapeutic unit to allow for precise and accurate conversation. Language is the primary tool for syntaxic communication. Non-verbal tools support the process and can be reference with language clearly but not directly communicated (qualia)
AI Potential Space AIPS, AI Self Evidencing Space AISES
Define space being explored — neither interoceptive nor exteroceptive — psycheceptive? Psychoactive, Psychoception, Psychotherapeutic manifold, n-dimensional developmental space; correlate with Bion theory of thinking but clarify that this model is inclusive of many modes of function and not inherently psychoanalytic
Markov Blanket selection: hierarchical, spontaneous vs planned, unity/dyad/group, individuals, sub groups — self-states, pairings of self-states, therapist within-therapist outside, sub dynamics t-p part inside — t-p part outside
Tracking Measure Candidates: Sense of Coherence, brain imaging, behavioral measures eg digital phenotyping,…
Phase 1: Capture — Engagement, Evaluation and Assessment; Initial Planning; Coupling Initiated
- Initiate engagement
- Incorporate conventional approaches for assessment
- AIT specific: determine utility of explicit empirical guided self-appraisal of patient’s subject experience in different states eg DMN vs task, how well-developed is baseline self-governance and self and world modeling? Where are areas of best and what areas are challenged in self and world modeling? How can we operationalize measuring self-world model fit (SWMF)? Quick rating and more comprehensive?
- Self-Classification
- Self-simulation Effectiveness, conscious recursive potential, current and max (if optimized), determinants
- Evaluate structure of personality, degree of unitary vs. plural self-states, theory of structural dissociation, degree of fragementation and dysconnectivity, etc., Incorporate multiplicity/pluralistic models e.g. Structural Theory of Dissociation, Brombergian relationship psychoanalysis (Standing in the Spaces)
- Different types and degrees of self-reflective function are useful in different contexts and time scales. Knowing which to use is key. Self-monitoring efficiency.
- Vocabulary level 1, 2, 3 (see below): e.g. Active Inference, Free Energy Principle, Evidencing, Self-Evidencing, Model, Policy, complex systems terms/concepts, uncertainty, surprise, complexity, accuracy, prior, posterior, Bayesian mentation, Markov(ian) process, Precision, Accuracy, Model Updating, Plasticity, etc.
- Develop graphical tools and metaphors, consider Feynman Diagram-like tool.
- Use AIT-specific assessments to determine problem areas including: model issues, modelling issues, updating issues, perceptual bias, affecting updating, and related (see below) using Markov Functions
- Evaluate individual to develop personalized approaches tailored to their level of understanding and preferred modes of communication, for example: to determine if they can make use of more technical information or if metaphorical approachs will be more effective, to evaluate communication style and match with therapist
- What space is being explored together? Psycheceptive space? A la Bion psychoanalytic function of the personality, reference Bion theory of thinking
- Determine initial set of hierarchical goals for therapy, subject to revision, both in terms of best guess order of operations as well as scale over context and time.
Phase 2: Initial Evidencing/Exploration/Experimentation — Greater Model Uncertainty — Lower Mastery
- Select top 1–3 approaches and implement with patient
- Track response using Active Inference-based tracking
- Update approach based on initial outcomes using technical formulations (whether explicit or implicit based on patient optimized approach)
- Refine diagnostic from Phase 1 to increase precision and accuracy of overall therapeutic model
- Characterized by “phase transitions” between more and less structured domains for example structured inquire based on active inference principles and processes, and more fluid psychoanalysis-like free-form stages of open exploration and self-evidencing. The time scale may vary e.g. longer periods of each, determine whether to structure sequencing or let it be free-form (or both).
Phase 3: Working Phase — Increased Baseline Model Certainty — Greater Mastery/Proficiency
- Consolidate understanding from initial phases with the patient
- Develop plan of action and set goals. This may be more or less open-ended, depending on the patient’s presentation and needs, on a continuum from CBT-like to Psychodynamic-like, with various tools and practices derived from related models
- Track progress and make small adjustments as needed
- Monitor for need to return to earlier phase to re-assess e.g. when developmental transitions and/or life changes require AIT model updating
Phase 4: Release — Consolidated/Stable/Resilient Updated Meta-Models, Models and Policies, Coupling Ends
- Anticipate stopping therapy and assess for optimal process e.g. taper off versus pick a date and stop (traditional termination)
- Review and consolidate prior work, including self-monitoring processes and changes to model updating and policy making meta-cognitive / reflective function
- Determine need for ongoing monitoring and related parameters
Phase 5: Post-termination internalization and tracking, potential return to therapy
· Monitor for recurrence based on pre-determinted parameters
· Define unpredicatables and have explicit plan
· Develop structured monitoring if appropriate
Illustrations and Visual Models and Metaphors
Final Goals and Metaphor:
AIT aims to help patients achieve optimal model complexity, adaptive surprise tolerance, efficient updating, emotional fluency, creative adaptability, systemic integration, sustainable growth, AI-human synergy, optimized neuroplasticity, and de-canalization of maladaptive patterns.
The mind is viewed as an adaptive ecosystem within the larger biosphere of society and culture. AIT helps patients become skilled ecologists and climate scientists of their internal and external worlds, using principles like TEMP to introduce adaptive “climate variations.” AI serves as advanced modeling technology in this complex neural-social-environmental system.
This comprehensive AIT model provides a flexible framework for understanding and treating a wide range of psychological issues, emphasizing the integration of active inference principles, neuroplasticity concepts, systems thinking, and technology in promoting mental health and personal growth.
Interview with Karl Friston, Doorknob Comments Podcast
AIT Model Notes
1. Foundational Principles:
AIT views the mind as a dynamic, adaptive system engaged in continuous modeling of self, others, and the environment, grounded in Bayesian mechanics, active inference principles, and neuroplasticity.
NB the specific pathological categories discussed below are placeholders and require elaboration.
Key concepts:
- Predictive Processing and Active Inference
- Embodied Cognition
- Social and Environmental Modeling
- Emotional Intelligence and Creative Inference
- Cultural and Systemic Influences
- Multiple Forms of Neuroplasticity
- Canalization and De-canalization
- Annealing in Neural Networks
- Temperature or Entropy Mediated Plasticity (TEMP)
2. Assessment, Education, and Intervention Framework:
- Comprehensive assessment of cognitive, emotional, social, and neuroplastic factors
- Patient education on active inference, neuroplasticity, and systems thinking
- Interventions targeting model updating, surprise tolerance, and adaptive neuroplasticity
3. Planning, Support, and Continuous Adaptation:
- Multi-horizon planning (short, mid, and long-term)
- Emphasis on supportive environments and relationships
- Ongoing assessment and adaptation of treatment
4. AI Integration:
AI tools enhance modeling, emotional regulation, and neuroplasticity promotion.
References and Additional Notes
Psychopathology Modeling and Interventions:
PSTD: Reframing PTSD for computational psychiatry with the active inference framework
Computational psychiatry: the brain as a phantastic organ
REBUS and the Anarchic Brain: Toward a Unified Model of the Brain Action of Psychedelics
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7280559/
https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2019.00592/full
https://www.mdpi.com/1099-4300/26/4/343
Process Model Citations:
[1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9163786/
[2] https://www.frontiersin.org/journals/human-neuroscience/articles/10.3389/fnhum.2022.955558/full
[3] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9280207/
[4] https://psycnet.apa.org/record/2023-28272-001
[5] https://www.nature.com/articles/s41598-021-89047-0
Friston and Collaborators:
- Friston, K. J., Stephan, K. E., Montague, R., & Dolan, R. J. (2014). Computational psychiatry: the brain as a phantastic organ. The Lancet Psychiatry, 1(2), 148-158.
- Carhart-Harris, R. L., & Friston, K. J. (2019). REBUS and the anarchic brain: toward a unified model of the brain action of psychedelics. Pharmacological Reviews, 71(3), 316-344.
- Parr, T., & Friston, K. J. (2017). Working memory, attention, and salience in active inference. Scientific Reports, 7(1), 1-21.
- Ramstead, M. J., Badcock, P. B., & Friston, K. J. (2018). Answering Schrödinger's question: A free-energy formulation. Physics of Life Reviews, 24, 1-16.
Active Inference, Free Energy, and Psychotherapy:
Arousal coherence, uncertainty, and well-being: an active inference account Hannah Biddell, Mark Solms, Heleen Slagter, Ruben Laukkonen https://academic.oup.com/nc/article/2024/1/niae011/7631817 Neuroscience of Consciousness, 2024
Holmes, J., & Nolte, T. (2019). "Surprise" and the Bayesian brain: Implications for psychotherapy theory and practice. Frontiers in Psychology, 10, 592.
Holmes J. Friston’s free energy principle: new life for psychoanalysis? BJPsych Bull. 2022 Jun;46(3):164–168. doi: 10.1192/bjb.2021.6. PMID: 33597069; PMCID: PMC9345684.
- Wilkinson, S., Deane, G., Suarez, K., & Rushton, J. A. (2019). Active inference and epistemic value in psychiatry. In Computational Psychiatry (pp. 27-52). Academic Press.
- Constant, A., Ramstead, M. J., Veissière, S. P., & Friston, K. (2019). Regimes of expectations: An active inference model of social conformity and human decision making. Frontiers in Psychology, 10, 679.
- Kirchhoff, M., Parr, T., Palacios, E., Friston, K., & Kiverstein, J. (2018). The Markov blankets of life: autonomy, active inference and the free energy principle. Journal of The Royal Society Interface, 15(138), 20170792.
- Seth, A. K., & Friston, K. J. (2016). Active interoceptive inference and the emotional brain. Philosophical Transactions of the Royal Society B: Biological Sciences, 371(1708), 20160007.
- Pezzulo, G., Maisto, D., Barca, L., & Van den Bergh, O. (2019). Symptom perception from a predictive processing perspective. Clinical Psychology in Europe, 1(4), 1-14.
- Fabry, R. E. (2020). Into the dark room: A predictive processing account of major depressive disorder. Phenomenology and the Cognitive Sciences, 19(4), 685-704.
- Allen, M., & Friston, K. J. (2018). From cognitivism to autopoiesis: towards a computational framework for the embodied mind. Synthese, 195(6), 2459-2482.
- Linson, A., Parr, T., & Friston, K. J. (2020). Active inference, stressors, and psychological trauma: A neuroethological model of (mal) adaptive explore-exploit dynamics in ecological context. Behavioural Brain Research, 380, 112421.
- Chekroud, A. M. (2015). Unifying treatments for depression: an application of the free energy principle. Frontiers in Psychology, 6, 153.
Foundational Principles:
a) Predictive Processing and Active Inference:
- Friston, K. (2010). The free-energy principle: a unified brain theory? Nature Reviews Neuroscience, 11(2), 127-138.
- Clark, A. (2013). Whatever next? Predictive brains, situated agents, and the future of cognitive science. Behavioral and Brain Sciences, 36(3), 181-204.
b) Embodied Cognition:
- Varela, F. J., Thompson, E., & Rosch, E. (2016). The embodied mind: Cognitive science and human experience. MIT press.
c) Psychoanalytic Foundations:
- Freud, S. (1923). The ego and the id. Standard Edition, 19, 1-66.
- Galatzer-Levy, R. M. (2017). Nonlinear Psychoanalysis: Notes from Forty Years of Chaos and Complexity Theory. Routledge
d) Interpersonal Relational Psychoanalysis:
- Mitchell, S. A. (1988). Relational concepts in psychoanalysis: An integration. Harvard University Press.
- Benjamin, J. (2017). Beyond doer and done to: Recognition theory, intersubjectivity and the third. Routledge.
e) Trauma Theory and Dissociation:
- Herman, J. L. (1992). Trauma and recovery: The aftermath of violence--from domestic abuse to political terror. Basic Books.
- Van der Hart, O., Nijenhuis, E. R., & Steele, K. (2006). The haunted self: Structural dissociation and the treatment of chronic traumatization. W. W. Norton & Company.
f) Neuroplasticity:
- Doidge, N. (2007). The brain that changes itself: Stories of personal triumph from the frontiers of brain science. Penguin.
g) Emotional Intelligence and Creative Inference:
- Goleman, D. (1995). Emotional intelligence. Bantam Books.
Assessment, Education, and Intervention Framework:
- McWilliams, N. (2011). Psychoanalytic diagnosis: Understanding personality structure in the clinical process. Guilford Press.
- Stern, D. N. (2004). The present moment in psychotherapy and everyday life. W. W. Norton & Company.
3. Planning, Support, and Continuous Adaptation:
- Fonagy, P., Gergely, G., Jurist, E. L., & Target, M. (2002). Affect regulation, mentalization, and the development of the self. Other Press.
- Wallin, D. J. (2007). Attachment in psychotherapy. Guilford Press.
AI Integration:
- Luxton, D. D. (2014). Artificial intelligence in psychological practice: Current and future applications and implications. Professional Psychology: Research and Practice, 45(5), 332-339.
Psychopathology Modeling and Interventions — Rough Notes:
a) Major Depressive Disorder:
- Blatt, S. J. (2004). Experiences of depression: Theoretical, clinical, and research perspectives. American Psychological Association.
b) Anxiety Disorders:
- Wachtel, P. L. (2008). Relational theory and the practice of psychotherapy. Guilford Press.
c) Bipolar Disorder:
- Goodwin, F. K., & Jamison, K. R. (2007). Manic-depressive illness: Bipolar disorders and recurrent depression. Oxford University Press.
d) Schizophrenia and Psychosis:
- Lysaker, P. H., & Dimaggio, G. (2014). Metacognitive capacities for reflection in schizophrenia: Implications for developing treatments. Schizophrenia Bulletin, 40(3), 487-491.
e) Obsessive-Compulsive Disorder:
- Gabbard, G. O. (2001). Psychoanalytically informed approaches to the treatment of obsessive-compulsive disorder. Psychoanalytic Inquiry, 21(2), 208-221.
f) Post-Traumatic Stress Disorder and Complex Trauma:
- Van der Kolk, B. A. (2014). The body keeps the score: Brain, mind, and body in the healing of trauma. Viking.
- Courtois, C. A., & Ford, J. D. (Eds.). (2009). Treating complex traumatic stress disorders: An evidence-based guide. Guilford Press.
g) Dissociative Disorders:
- Putnam, F. W. (1989). Diagnosis and treatment of multiple personality disorder. Guilford Press.
- Howell, E. F. (2011). Understanding and treating dissociative identity disorder: A relational approach. Routledge.
h) Personality Disorders:
- Kernberg, O. F. (1984). Severe personality disorders: Psychotherapeutic strategies. Yale University Press.
- Linehan, M. M. (1993). Cognitive-behavioral treatment of borderline personality disorder. Guilford Press.
i) Substance Use Disorders:
- Khantzian, E. J. (1997). The self-medication hypothesis of substance use disorders: A reconsideration and recent applications. Harvard Review of Psychiatry, 4(5), 231-244.
j) Attachment-Based Approaches:
- Bowlby, J. (1988). A secure base: Parent-child attachment and healthy human development. Basic Books.
- Schore, A. N. (2003). Affect regulation and the repair of the self. W. W. Norton & Company.
k) Interpersonal Neurobiology:
- Siegel, D. J. (2012). The developing mind: How relationships and the brain interact to shape who we are. Guilford Press.
l) Mindfulness and Psychotherapy:
- Germer, C. K., Siegel, R. D., & Fulton, P. R. (Eds.). (2005). Mindfulness and psychotherapy. Guilford Press.
m) Neuropsychoanalysis:
- Solms, M., & Turnbull, O. (2002). The brain and the inner world: An introduction to the neuroscience of subjective experience. Other Press.
This revised reference list now incorporates key psychoanalytic thinkers, including Freud and Levy, as well as significant contributors to interpersonal relational psychoanalysis, trauma theory, and the treatment of dissociative disorders. It provides a more comprehensive foundation for the Active Inference Therapy model, bridging traditional psychoanalytic thought with contemporary neuroscience and cognitive theories.
Here are specific models and recommendations for key psychiatric conditions, including examples of change processes and alternatives for obstacles:
a) Major Depressive Disorder (MDD):
Model: MDD is viewed as a prior expectation of negative outcomes, leading to reduced exploratory behavior and anhedonia.
Intervention: Gradual exposure to positive experiences, combined with TEMP-based variability training.
Example: A patient starts with small, achievable daily goals (e.g., a 5-minute walk). Each successful experience updates their predictive model, gradually shifting expectations. TEMP principles introduce variability in activities to prevent rigid thinking.
Obstacle: Patient struggles with motivation.
Alternative: Use AI-assisted virtual reality experiences to simulate positive outcomes, building motivation for real-world engagement.
b) Generalized Anxiety Disorder (GAD):
Model: GAD is seen as hyperactive prediction error signaling, resulting in excessive threat detection and worry.
Intervention: Recalibrating the balance between priors and sensory evidence through mindfulness and gradual exposure.
Example: Patient learns mindfulness techniques to enhance interoceptive awareness, recognizing the difference between predicted and actual bodily sensations. Gradual exposure to worry-inducing situations helps update threat-related priors.
Obstacle: Patient experiences panic during exposure.
Alternative: Use AI biofeedback tools to help patient recognize and regulate physiological responses during graduated exposure.
c) Bipolar Disorder:
Model: Conceptualized as cyclical fluctuations in predictive models, with manic states representing overly optimistic priors and depressive states overly pessimistic ones.
Intervention: Stabilizing these fluctuations through consistent model updating and rhythm regulation.
Example: Patient uses mood tracking apps to identify early signs of mood shifts. They engage in regular sleep-wake cycle maintenance and social rhythm therapy to stabilize their predictive models.
Obstacle: Patient resists mood stabilization, valuing manic productivity.
Alternative: Explore creativity-enhancing techniques that don’t rely on mood fluctuations, demonstrating stable productivity possibilities.
d) Schizophrenia:
Model: Understood as a breakdown in the balance between prior beliefs and sensory evidence, leading to delusions and hallucinations.
Intervention: Strengthening reality-testing abilities and promoting adaptive neuroplasticity.
Example: Patient engages in cognitive remediation exercises to enhance processing of sensory information. They practice distinguishing between internal and external stimuli through grounding techniques.
Obstacle: Persistent delusions resist change.
Alternative: Instead of directly challenging delusions, focus on enhancing overall cognitive flexibility and social functioning.
e) Obsessive-Compulsive Disorder (OCD):
Model: Viewed as overly rigid predictive models leading to compulsive behaviors.
Intervention: Exposure and response prevention techniques, explained through Bayesian updating principles.
Example: Patient gradually exposes themselves to anxiety-provoking stimuli without performing compulsions, updating their predictive models about threat and control.
Obstacle: Patient unable to tolerate exposure anxiety.
Alternative: Use virtual reality for initial exposures, gradually transitioning to real-life situations.
f) Post-Traumatic Stress Disorder (PTSD):
Model: Seen as maladaptive neuroplasticity in fear circuits and overgeneralized threat detection.
Intervention: Promoting new, adaptive neuroplastic changes and updating trauma-related priors.
Example: Patient undergoes EMDR therapy, reconsolidating traumatic memories and updating associated predictive models. They practice grounding techniques to enhance present-moment awareness.
Obstacle: Patient experiences flashbacks during treatment.
Alternative: Use neurofeedback to help patient recognize and regulate autonomic arousal during trauma processing.
g) Alcohol Use Disorder:
Model: Understood as substance-induced neuroplasticity leading to maladaptive reward processing and inflexible behavior patterns.
Intervention: Restoring homeostatic plasticity, updating alcohol-related priors, and enhancing behavioral flexibility.
Example: Patient engages in cue exposure therapy, updating their predictive models about alcohol’s effects and their ability to resist cravings. They learn and practice alternative coping strategies to replace drinking behaviors.
Obstacle: Patient experiences intense cravings in social situations.
Alternative: Use role-playing and virtual reality simulations to practice social scenarios without alcohol, gradually transitioning to real-world situations.
h) Borderline Personality Disorder (BPD):
Model: Seen as highly unstable predictive models of self and others, with impaired emotion regulation.
Intervention: Stabilizing these models, enhancing emotional regulation, and promoting interpersonal effectiveness.
Example: Patient learns dialectical behavior therapy skills to enhance emotion regulation and interpersonal effectiveness. They practice mentalization techniques to improve their modeling of others’ mental states.
Obstacle: Patient struggles with maintaining therapeutic alliance.
Alternative: Use AI-assisted coaching between sessions to provide consistent support and reinforce therapeutic gains.
i) Attention-Deficit/Hyperactivity Disorder (ADHD):
Model: Viewed as difficulties in predictive processing and attentional allocation.
Intervention: Enhancing predictive capabilities, attentional control, and executive functioning.
Example: Patient engages in cognitive training exercises to improve working memory and attention. They learn to structure their environment to support focus and task completion.
Obstacle: Patient forgets to implement strategies.
Alternative: Use smartphone apps with AI-powered reminders and task management to support daily functioning.
j) Autism Spectrum Disorder (ASD):
Model: Understood as atypical predictive processing, particularly in social domains.
Intervention: Enhancing social prediction skills and promoting adaptive neuroplasticity in social cognition.
Example: Patient participates in social skills training, using video modeling and role-playing to enhance their understanding of social cues and norms. They practice theory of mind exercises to improve social prediction.
Obstacle: Patient experiences sensory overload in social situations.
Alternative: Use virtual reality social simulations with controllable sensory input to practice social skills in a manageable environment.
k) Post Traumatic and Dissociative Disorders
