A path to Reconceptualizing Psychological Concepts through Active Inference

Introduction and framing

Psychological concepts such as belief, the unconscious, and defense mechanisms have been central to our understanding of how people think, feel, and behave. These ideas form the foundation of many traditional psychological models, offering a means to explore the complex interactions of mental processes and behaviors. However, while these frameworks have been influential, recent advancements in neuroscience, particularly in the understanding of brain function, provide new dimensions for comprehending these long-standing ideas. By integrating these developments, we can enrich our psychological models with a biologically grounded perspective that captures not only the abstract but also the embodied aspects of human cognition.

The Free Energy Principle (FEP) and active inference have emerged as powerful frameworks for understanding the mind and body as predictive systems. These concepts do not negate traditional psychological theories; rather, they provide a complementary approach that builds on them, offering a deeper and more dynamic understanding of mental and bodily processes. At its core, the FEP posits that living systems, including humans, strive to minimize surprise or prediction error by continuously updating their internal models of the world. This minimizes free energy, a theoretical construct reflecting the discrepancy between expected and actual sensory input. The brain and body, in this view, function as systems constantly engaged in predicting and adapting to their environments to maintain homeostasis.

Active inference, a key component of the FEP, elaborates on how this predictive process unfolds. It suggests that the body does not passively respond to stimuli but actively seeks to confirm its predictions through sensory and motor actions. This process of prediction and correction is not limited to the brain; it is deeply embodied, with the body playing a crucial role in refining these predictions through action. This resonates with embodied cognitive science, particularly the 4E (embodied, embedded, enacted, and extended) approach, which emphasizes that cognition cannot be understood in isolation from the physical body and its interactions with the environment.

By framing the mind and body as an integrated predictive system, the FEP provides a biologically grounded theory that can enhance our understanding of traditional psychological concepts. For example, defense mechanisms in psychodynamic theory, such as repression or denial, could be reinterpreted as cognitive strategies to minimize prediction error. The unconscious, rather than being merely a repository of repressed desires, might be seen as the backdrop of unarticulated predictions that shape our conscious experience without explicit awareness. This reinterpretation enriches traditional psychological theories by providing a mechanistic understanding of the processes that drive mental and bodily regulation.

The practical implications of FEP and active inference extend beyond theory into computational modeling and clinical practice. The mathematical framework underpinning these concepts allows researchers to simulate how the brain and body minimize prediction error and adapt to their environments. These simulations can be used to understand a range of psychological phenomena, from anxiety and depression to more complex disorders like schizophrenia. Furthermore, wearable bio-data technologies and immersive tools such as Virtual Reality (VR) and Augmented Reality (AR) offer exciting opportunities to apply FEP principles in therapeutic settings. By manipulating sensory inputs in controlled ways, these technologies can help individuals recalibrate their internal models, reduce prediction errors, and enhance emotional regulation and psychological well-being.

This dynamic perspective not only reshapes our understanding of mental processes but also positions the FEP as a bridge between traditional psychological theories and cutting-edge neuroscientific research. It acknowledges that while traditional theories have provided profound insights into the human condition, their integration with biologically grounded frameworks like FEP can deepen our understanding of how the mind, body, and environment interact to shape human experience. In doing so, this approach opens up new avenues for both research and clinical applications, enhancing the precision and effectiveness of psychological interventions.

Reimagining Psychological Concepts

In recent years, the integration of neuroscience with psychodynamic theory has opened up new avenues for understanding the mechanisms of therapeutic change. One such framework, the Free Energy Principle (FEP), developed by Karl Friston, provides a novel lens through which to explore how humans, as predictive systems, constantly strive to minimize surprise or free energy by continuously updating their internal models of the world. Jeremy Holmes, in his exploration of the FEP within psychoanalytic psychotherapy, highlights how this principle aligns with key psychodynamic concepts such as transference, self-reflection, and the therapeutic alliance. Holmes suggests that therapy itself can be viewed as a process of active inference, where patients gradually revise maladaptive internal models based on their interactions within the therapeutic setting. By reinterpreting transference as the projection of outdated predictive models onto the therapist, and self-reflection as a mechanism for reducing prediction errors, this framework bridges traditional psychoanalytic concepts with a biologically grounded understanding of the mind and body. This integration enriches both fields, offering a deeper insight into the processes that drive emotional regulation, behavioral change, and therapeutic success.

?

Belief: The Body’s Predictive Framework

In traditional psychology, beliefs are understood as mental convictions or strong opinions that influence how we perceive the world, make decisions, and interact with others. For example, if someone believes they are good at their job, this belief will influence their confidence, the way they approach tasks, and how they respond to challenges. These beliefs are typically considered stable parts of our minds—once we form a belief, it tends to stick unless something significant changes our mind.

Active inference offers a new perspective on beliefs, viewing them not as fixed opinions but as part of a dynamic, ongoing process. In this framework, beliefs are probabilistic expectations that the brain and body use to predict future events. These predictions are continuously tested and updated based on new information. For example, if someone expects a friendly response and receives a smile, their belief is reinforced, encouraging further interaction. However, if they expect rejection and receive an ambiguous response, their belief may lead them to withdraw or react negatively.

The mathematical framework provided by the FEP allows these belief systems to be modeled computationally. Through algorithms that simulate the body's predictive processes, researchers can study how beliefs are formed, maintained, and altered. This computational modeling is further enriched by incorporating bio-data from wearables, which monitor physiological states in real-time, providing insights into how the body's predictions align with external reality.

Moreover, VR and AR technologies can be utilized to create controlled environments where these beliefs can be tested and modified. Future studies could use experimental paradigms, such as social simulations or virtual reality, to test how individuals with depression process social information and how their generative models change in response to new social experiences. For instance, a VR scenario could simulate public speaking to help an individual confront and reassess their fear of public speaking. By gradually increasing the complexity and realism of the scenario, and using biofeedback from wearables to monitor the user's physiological responses, therapists can guide the individual through a process of belief recalibration in a safe and controlled setting.

This perspective on beliefs emphasizes that they are not fixed but are continuously shaped by our experiences and interactions with the world. The body is always working to minimize prediction errors, so beliefs are constantly being adjusted to keep our mental models of the world accurate and useful.

?

Consciousness as Predictive Regulation

Mark Solms’ interpretation of consciousness through the lens of the Free Energy Principle reframes it as a process rooted in biological imperatives. He emphasis on affect shifts the traditional cognitive-centric focus of consciousness studies toward a more embodied understanding. In this way, consciousness is fundamentally tied to the body's efforts to stay ahead of uncertainty, with affective states representing the organism's real-time assessment of its ability to maintain homeostasis. As the brain and body work to reduce free energy, consciousness serves as the subjective experience of navigating and correcting these prediction errors in the moment.

He argues that affect, the core component of feeling, is intrinsic to consciousness because it directly relates to the organism’s drive to survive. This perspective is supported by evidence from neuroscience showing that affective experiences, such as fear, hunger, or pleasure, are mediated by brain structures in the brainstem and subcortical areas—regions evolutionarily older than the cerebral cortex. These structures are involved in maintaining homeostasis and regulating the autonomic nervous system, indicating that consciousness originates from a primal level of biological functioning. By placing affect at the center of conscious experience, Solms suggests that the subjective quality of consciousness—why it "feels like something" to be conscious—is a direct result of the body's need to minimize uncertainty in relation to survival. The more an organism successfully predicts and corrects deviations from expected states, the more consciousness is experienced as stable and integrated.

?

Emotion: Interoceptive Prediction and Adaptive Behavior

Emotions have traditionally been understood as complex psychological states that involve a mix of subjective feelings (like happiness or anger), physiological responses (like a racing heart or sweating), and behavioral expressions (like smiling or frowning). For example, when you feel afraid, you might notice your heart beating faster, a knot in your stomach, and the urge to run away. These components of emotion are usually seen as reactions to things that happen to us—external events that trigger internal responses.

Active inference offers a deeper understanding of emotions by framing them as part of the body’s predictive processes. According to this view, emotions are not just reactions to what happens around us but are predictions that our body makes about what will happen internally. This process is known as interoception, which refers to the body’s ability to monitor and predict its internal states.

For instance, when you feel anxious, your body is predicting that something bad might happen, which triggers a series of physiological responses: your heart rate increases, your muscles tense, and your breathing might become shallow. These physiological changes are the body’s way of preparing to respond to a potential threat. If the predicted threat doesn’t materialize, the body adjusts its model, and the anxiety may subside.

Computational modeling can simulate these interoceptive predictions, allowing researchers to predict how different emotional states might arise under various conditions. By using bio-data from wearables, such as heart rate monitors or skin conductance sensors, therapists can gain real-time insights into the physiological basis of these emotional states. This data can then be used to tailor therapeutic interventions, helping individuals learn to better regulate their emotions by adjusting their interoceptive predictions.

This way of thinking about emotions as predictions has important implications for understanding and treating emotional disorders. By viewing emotions as part of a dynamic process of prediction and adjustment, we can better understand how they work and develop more effective ways to manage them.

Self: The Minimal Phenomenal Selfhood

The concept of self in psychology traditionally refers to the individual's perception of their own identity and existence. This sense of self has often been treated as a core part of one's personality and consciousness—a stable "I" that persists over time. People usually think of the self as something that is uniquely theirs, a consistent presence that defines who they are across different situations.

However, within the active inference framework, the self is reinterpreted as a dynamic, ongoing process rather than a fixed entity. This idea aligns with the concept of minimal phenomenal selfhood (MPS), which suggests that the self is not a static thing but an experience that the body constantly creates and updates. According to the Free Energy Principle (FEP), the self is a model that the body generates to predict and make sense of its interactions with the world.

This model is continuously shaped by sensory input and bodily experiences. For example, when you move your hand, your body predicts what that movement will feel like and what it will look like. If everything matches the prediction, your sense of self as a person who can move and control their body is reinforced. But if something goes wrong—if the movement doesn’t go as planned—your body might adjust its model of the self to account for this discrepancy.

The FEP provides a mathematical framework for modeling the self as a set of hierarchical predictions that are constantly updated based on new sensory and interoceptive information. Wearables can track physiological states that contribute to this sense of self, such as body movement, heart rate, and respiration. This real-time data can inform computational models that simulate how the self is constructed and maintained, providing insights into conditions where this process might be disrupted, such as in depersonalization or schizophrenia.

Minimal Phenomenal Selfhood (MPS) refers to the immediate, pre-reflective sense of being someone. Unlike higher cognitive functions, MPS emerges from the organism's embodied interaction with the world, where the body plays a central role. It acts both as a physical entity and an enabler of self-conscious experience. This notion ties closely to the Free Energy Principle (FEP), which posits that biological organisms, including the human brain, function to minimize free energy or prediction errors. Through this process of continuously updating internal models based on sensory inputs, the organism maintains homeostasis and cognitive coherence, ensuring its survival and well-being.

Embodied cognition, a key aspect of MPS, underscores the body’s integral role in cognition. The body serves as the vehicle for interaction with the world, with the body predicting and integrating multisensory information to sustain a coherent sense of self. Hierarchical generative models and predictive coding explain how the brain processes sensory data at different levels of abstraction, integrating both interoceptive (internal) and exteroceptive (external) inputs. This results in the cohesive self-model responsible for the experience of "mineness," or the feeling that bodily experiences belong to the self. Agency, an essential component of MPS, reflects the organism's ability to act and adjust behavior through active inference, continuously refining its predictions.

A vivid example of the body’s role in shaping the sense of self is illustrated in the rubber hand illusion. In this experiment, a participant watches a rubber hand being stroked while their real hand (which is hidden from view) is stroked simultaneously. Over time, the participant begins to feel as though the rubber hand is their own. This illusion highlights how the brain integrates multisensory information (in this case, visual and tactile input) to update its self-model and create a sense of bodily ownership. It also shows the flexibility and dynamic nature of the self-model, which can be manipulated by altering sensory inputs.

Applications of these concepts are evident in phenomena like the rubber hand illusion, where the body flexibly manipulates self-representations based on altered sensory input, highlighting the dynamic nature of the self-model.

This way of thinking about the self emphasizes that it is not a fixed, unchanging entity. Instead, it is something that is constantly being constructed by the body as it processes information from its environment and its internal states. This dynamic view of the self highlights the deep connection between our bodies, our sensory experiences, and our sense of identity. It suggests that who we are is not just something we think about abstractly but is something that is lived and felt through our bodies in every moment.

A practical application of active inference is in understanding self-esteem. Traditionally viewed as a stable trait reflecting an individual's sense of self-worth, self-esteem can be seen more dynamically through the lens of active inference. In this framework, self-esteem functions as a generative model that continuously updates based on social feedback to minimize the gap between expected and actual experiences. Positive feedback reinforces self-esteem, aligning an individual's internal sense of worth with external validation, while negative interactions or perceived rejection can prompt a downward revision, leading to uncertainty and emotional distress. The active inference model also distinguishes between state self-esteem, which fluctuates in response to specific events, and trait self-esteem, a more enduring sense of self-worth. This distinction highlights how short-term social experiences influence long-term self-perception. Importantly, this process is embodied—individuals not only interpret external feedback but also integrate internal emotional signals, like pride or shame, to shape their self-concept.

Reimagining Transference through Active Inference

Transference has long been a cornerstone of psychoanalytic theory, traditionally understood as the unconscious redirection of feelings from one person (often a significant figure from the past) onto another, such as the therapist. This process has been central to understanding how patients' early life experiences influence their present behaviors and emotional responses in therapy. However, the lens of active inference and the Free Energy Principle (FEP) allows us to reinterpret transference as part of the body and brain's predictive model, offering deeper insight into its mechanics.

From the perspective of active inference, transference can be viewed as a set of deeply ingrained prior predictions about interpersonal relationships. These priors, formed in response to early attachment experiences, shape how individuals expect others to behave and interact with them. In the context of therapy, transference occurs when these outdated predictions about relationships surface and are projected onto the therapist. In this way, transference becomes a mechanism for exploring the client’s internal predictive models and the errors that arise when these models no longer fit current interpersonal dynamics.

The therapeutic setting provides an ideal environment for recalibrating these priors. The therapist’s neutral or ambiguous stance acts as a "safe" yet uncertain context, which allows the patient’s predictions to be challenged and corrected. When the therapist does not respond according to the patient's expectations (i.e., the prior), this creates prediction error—a discrepancy between what is anticipated and what is experienced. This prediction error, in turn, opens up a space for active inference, where the patient can begin to revise and update their internal models about relationships. The process of exploring these prediction errors in therapy mirrors the Bayesian updating mechanism central to the Free Energy Principle: clients learn to minimize surprise by refining their relational expectations to better match reality.

?

Self-Reflection and Active Inference

Self-reflection, another key component of many therapeutic modalities, can also be reconceptualized through the framework of active inference. Traditionally, self-reflection is understood as the process by which individuals introspect, analyze their thoughts and emotions, and gain insight into their mental and emotional states. While this process is often framed as purely cognitive, the Free Energy Principle allows us to view self-reflection as a dynamic, embodied process aimed at reducing uncertainty and maintaining internal coherence.

According to active inference, self-reflection involves continuously updating one's internal model of the self to minimize prediction error. This model is shaped by sensory, interoceptive, and affective feedback, which informs the individual about their state of being. When discrepancies arise between the individual’s internal model of themselves (their beliefs, values, or emotional state) and external feedback (social or environmental cues), self-reflection facilitates the process of reconciling these differences. It enables the person to modify their predictions and adjust their behavior accordingly, thus reducing uncertainty and restoring a sense of coherence.

For example, in moments of emotional distress, an individual may experience a mismatch between their predicted emotional response and their actual emotional experience. Self-reflection allows them to examine this prediction error and reassess their internal model. By recognizing maladaptive patterns of thought or behavior, they can engage in active inference, refining their self-concept to better align with reality. This process is not just an intellectual exercise but a deeply embodied one, where affective and sensory signals play a crucial role in shaping the individual’s evolving sense of self.

In therapy, fostering self-reflection is an essential tool for helping patients revise rigid or inaccurate models of the self. The therapeutic relationship, by providing a space where feedback is offered in a supportive yet challenging manner, can facilitate the reduction of free energy. As clients reflect on their thoughts, emotions, and behaviors within this context, they are given the opportunity to engage in active inference, gradually refining their self-model in ways that promote greater emotional and psychological flexibility.

?

Resilience, Allostatic Load, and Well-Being

The concept of resilience is often understood as an individual's capacity to maintain psychological and physiological stability in the face of adversity. It is closely related to the management of allostatic load, which refers to the cumulative physiological burden placed on the body by chronic stress. Through the framework of active inference, resilience can be reconceptualized as an adaptive process by which individuals minimize the uncertainty and prediction errors that arise in challenging environments, thereby maintaining homeostasis and well-being.

Allostasis is the process through which the body adjusts its physiological set points in response to stress, constantly recalibrating to maintain balance. Over time, repeated or chronic stress can lead to an increased allostatic load, causing wear and tear on bodily systems. Active inference, with its focus on prediction and error correction, provides a mechanism for understanding how the brain and body adapt to reduce this load. By continuously updating internal models to predict and respond to environmental demands, individuals can better manage their physiological states, avoiding the detrimental effects of chronic stress.

Resilience, within the active inference framework, is the ability to effectively manage and reduce allostatic load through adaptive cognitive and emotional regulation strategies. Individuals with high resilience are better equipped to monitor their well-being, update their beliefs about stressful situations, and take actions that minimize uncertainty and prediction errors. This process is essential for maintaining subjective well-being (SWB), which is closely linked to how efficiently one’s internal generative models predict and mitigate stressors in real-time.

For example, individuals who can flexibly adjust their predictions about stress, altering their expectations based on feedback from the environment, are less likely to experience the prolonged physiological strain that leads to high allostatic load. Positive emotions, which enhance flexibility in thinking and problem-solving, play a key role in this process by enabling resilient individuals to reassess and recalibrate their predictions, ultimately fostering emotional and physiological recovery from stress.

By integrating the concepts of allostasis and active inference, we gain a dynamic understanding of how resilience supports well-being. The brain’s constant striving to reduce prediction error allows individuals to mitigate the effects of stress and uncertainty, thereby maintaining a state of balance and minimizing the long-term consequences of allostatic load. This framework also opens new avenues for therapeutic interventions that aim to enhance resilience by helping individuals develop adaptive strategies for managing stress, recalibrating their internal models, and maintaining well-being over time.

?

Dreaming as Complexity Reduction and Emotional Processing

Dreaming offers a unique window into how the brain operates as a predictive system, attempting to minimize free energy even during sleep. From the perspective of the Free Energy Principle (FEP), dreaming can be understood as a process through which the brain resolves emotional conflicts, consolidates memories, and reduces computational complexity. During sleep, particularly in rapid eye movement (REM) sleep, the brain generates virtual environments—dreams—that simulate various scenarios, allowing it to rehearse and integrate emotionally charged experiences.

Within the active inference framework, dreams are not just random narratives but serve as an adaptive mechanism for resolving prediction errors. When the brain encounters unresolved emotional conflicts during the day, it uses the relative safety of sleep to process these conflicts in a less constrained environment. Dreams provide a form of "virtual reality," where individuals can explore alternative outcomes and modify their generative models without facing real-world consequences. This aligns with the idea of predictive coding, where the brain uses dreams to test and refine its internal models, reducing free energy by simulating and resolving potential future scenarios.

Furthermore, dreaming plays a critical role in emotional regulation. Traumatic or emotionally complex experiences generate high levels of free energy, creating emotional tension that the brain seeks to minimize. During REM sleep, the brain actively engages in synaptic pruning and memory consolidation, reducing unnecessary or conflicting connections formed during waking life. This process helps to restore balance by simplifying the brain's emotional and cognitive landscape. When this complexity reduction fails—due to trauma, disrupted sleep, or other factors—unresolved emotions may manifest as psychiatric symptoms like anxiety or intrusive thoughts. In this sense, failures in dreaming may contribute to the persistence of psychological disorders.

Thus, dreaming is a fundamental mechanism through which the body maintains emotional and cognitive homeostasis, reducing complexity and updating its predictive models to ensure optimal functioning. By understanding dreams through the FEP, we gain deeper insight into their role in psychological well-being and mental health.

?

Computational Phenotypes and Personality

Computational phenotypes can indeed be reconceptualized as personality concepts in psychology. Traditionally, personality has been understood as a relatively stable set of traits (e.g., the Big Five: openness, conscientiousness, extraversion, agreeableness, neuroticism), which influence how individuals perceive the world, respond to it, and interact with others. These traits have been treated as fixed or inherent qualities. However, the concept of computational phenotypes, grounded in the active inference framework, offers a more dynamic and computational model of personality, emphasizing the processes by which the body generates and updates predictions about the world to minimize prediction error.

Dynamic and Contextual Nature of Traits

In traditional models, personality traits are considered stable across time and situations. Computational phenotypes, however, suggest that personality is fluid, emerging from the ongoing predictive processes of the body. Instead of viewing personality traits as fixed qualities, individuals can be understood in terms of their unique predictive models and how they respond to uncertainty and environmental feedback. For example, a person traditionally labeled as "neurotic" might possess a computational phenotype characterized by a heightened sensitivity to uncertainty, leading to increased anxiety in unpredictable situations. Rather than being an immutable trait, this sensitivity can fluctuate based on context and can be modified through learning and experience, reconceptualizing neuroticism as a flexible, context-dependent process.

Learning Rates and precision of priors

Personality traits such as openness to experience and conscientiousness can be reinterpreted through the lens of precision of priors and learning rates within computational phenotypes, offering a more dynamic, process-oriented understanding of behavior. Precision of priors refers to the confidence or rigidity with which individuals hold their predictions about the world. For example, those high in openness may exhibit lower precision in their priors, allowing them to flexibly update beliefs and expectations when confronted with new information or experiences. In contrast, individuals with more rigid priors may struggle to integrate new data into their worldview, reflecting lower openness.

Similarly, learning rates provide another dimension to understanding personality traits related to adaptability and flexibility. Individuals with high learning rates are more responsive to new information, quickly adjusting their predictions in changing environments. For instance, they may excel in fast-paced social or work contexts, rapidly adapting to feedback. Conversely, those with slower learning rates may exhibit more introverted tendencies or prefer structured environments, where adaptation is gradual.

Sensitivity to Uncertainty

Sensitivity to uncertainty is a key component of computational phenotypes and can be directly related to the personality trait of neuroticism. In traditional psychology, neuroticism is characterized by a predisposition to experience negative emotions such as anxiety, anger, and depression. Computational phenotypes frame this as heightened sensitivity to uncertainty and lower tolerance for ambiguity, leading to an overestimation of potential threats and resulting in anxiety. This reconceptualization offers a more granular understanding of neuroticism, not as a fixed trait, but as a consequence of how the body processes and responds to uncertainty, which can vary across situations and be modulated through intervention.

Agency, Autonomy, and Control in Personality

The computational framework of agency, where individuals act to minimize prediction errors and maintain control over their environment, can be mapped onto personality concepts like locus of control and self-efficacy. Individuals with a strong sense of agency—those who feel they can influence outcomes and control their environment—may have high levels of self-efficacy and an internal locus of control in traditional personality theory. On the other hand, individuals with computational phenotypes that reflect lower confidence in reducing prediction errors might align with an external locus of control, leading to feelings of helplessness or reliance on external forces. This reconceptualization situates personality traits like locus of control within the dynamic processes of predictive regulation.

Interpersonal Dynamics and Personality

Active inference’s focus on social prediction and interpersonal dynamics provides a novel lens to understand traits like agreeableness and extraversion in social environments. Individuals with computational phenotypes that prioritize reducing social prediction error may be more adept at interpreting social cues and adjusting their behavior to fit social norms, aligning with high agreeableness. In contrast, those who struggle with social prediction may exhibit behaviors that reflect lower agreeableness, often resulting in social conflict or withdrawal.

Integration of Psychoanalytic and Neuroscientific Perspectives

The integration of psychoanalytic concepts with neuroscientific frameworks, particularly through the lens of active inference, provides a rich terrain for rethinking long-standing ideas in psychology.

Implications for Therapy and Practice

Integration of the Free Energy Principle with psychoanalysis offers profound implications for clinical practice, particularly in understanding mental disorders. Conditions like anxiety and depression can be understood as maladaptive prediction systems, where the body overestimates threats or underestimates positive outcomes, leading to chronic states of heightened free energy (i.e., uncertainty). Psychoanalytic concepts, such as repression or transference, can be reinterpreted through active inference as strategies the body uses to minimize overwhelming prediction errors. For instance, repression might be seen as the body's attempt to avoid prediction errors related to traumatic memories by keeping them out of conscious awareness. In therapy, the process of surfacing these unconscious patterns can be viewed as a recalibration of the body’s internal model, allowing the individual to update their predictions and reduce prediction errors. By applying active inference, clinicians can guide patients through therapeutic processes that help recalibrate their internal models of the world, reducing the emotional and physiological uncertainties that underlie psychological distress.

Therapeutic practices can move towards predictive coding therapies that target the recalibration of maladaptive priors. In other words, therapy can focus on helping the body adjust its predictions so that they better match reality. This approach is particularly useful for treating conditions like anxiety, depression, and trauma, where the body’s predictions about the world are often overly negative or fearful. For example, in anxiety disorders, the body might predict that certain situations are dangerous, even when they are not. This prediction leads to feelings of anxiety and avoidance behaviors, which reinforce the belief that the situation is dangerous. By gradually exposing someone to the situations they fear in a controlled way (a process known as exposure therapy), the body can learn that these situations are not as threatening as it once thought.

Interoceptive Technologies and Their Role in Enhancing Therapy

Interoception, the process by which the body monitors and predicts its internal states, is central to understanding how we experience emotions and maintain a sense of self. Advances in interoceptive technologies have opened new avenues for both research and clinical practice by allowing precise manipulation and assessment of these internal bodily signals. These technologies offer the potential to improve our understanding of how interoception contributes to mental health and to develop targeted interventions for conditions where interoceptive processing is dysfunctional.

These technologies can be used as diagnostic tools to assess the sensitivity and accuracy of a patient's interoceptive system. For instance, in treating anxiety, controlled exposure to interoceptive stimuli such as increased heart rate or breathlessness can help patients learn to regulate their physiological responses to stress. This can reduce the intensity of anxiety symptoms and improve overall emotional regulation.

Moreover, these technologies can facilitate the development of personalized treatment plans by providing real-time feedback on how a patient's body responds to different interventions. This allows for more precise and effective therapeutic approaches that address both the cognitive and embodied aspects of mental health.

?

Categories of Interoceptive Technologies

Interoceptive technologies can be broadly classified into three categories: artificial sensations, interoceptive illusions, and emotional augmentation. Each category provides different methods for intervening in the body’s internal processes, with potential applications in both diagnosis and treatment.

1. Artificial Sensations

This approach involves the direct, bottom-up modulation of interoceptive signals at the level of receptor cells. Techniques such as C-tactile stimulation and panicogenic hypercapnia exemplify how controlled stimulation of the autonomic nervous system can influence emotional and cognitive states. For instance, C-tactile stimulation, which involves gentle, slow stroking of the skin, can modulate the body’s interoceptive system by activating unmyelinated C-fibers that are connected to the insular cortex. This can lead to a calming effect, reducing feelings of social exclusion and modulating the autonomic parasympathetic response. Conversely, hypercapnia, induced by elevated CO2 levels, can provoke panic attacks by disrupting normal respiratory function, providing a powerful tool for both understanding and treating anxiety disorders.

2. Interoceptive Illusions

These involve the manipulation of interoceptive signals through contextual or sensory modifications that alter the body’s perception of its internal state. For example, false feedback about heart rate or respiration can create an illusion that alters the perception of effort or emotional intensity. Such illusions can be used to study how interoceptive predictions are formed and updated, and they offer potential therapeutic applications for conditions like panic disorder or depression by allowing patients to recalibrate their internal models of bodily states.

3. Emotional Augmentation

This technique combines artificial sensations with interoceptive illusions and contextual cues to generate specific emotional states. For example, integrating exteroceptive stimuli, such as sounds or visuals associated with personal significance, with interoceptive manipulations can enhance the emotional experience. This approach has significant implications for exposure therapy, where the goal is to create controlled emotional responses that can help patients confront and process traumatic memories or phobias.

Integrating Bio-Data and Wearables

Wearable devices that track physiological signals—such as heart rate, skin conductance, respiratory patterns, and movement—are integral to the application of interoceptive technologies. These devices provide real-time bio-data that can be fed into computational models, allowing for continuous monitoring and adjustment of therapeutic interventions based on the patient’s current state.

For example, during exposure therapy, wearables can monitor how a patient’s heart rate and skin conductance change in response to anxiety-inducing stimuli. This data can help therapists assess the effectiveness of the intervention and make real-time adjustments to the treatment plan. By using bio-data, therapists can tailor their approaches to the individual’s specific physiological responses, leading to more personalized and effective therapy.

Additionally, wearables can be used outside of the clinical setting to help patients manage their conditions in real-time. For instance, a wearable device might alert a patient when their physiological signals indicate rising anxiety, prompting them to use a coping strategy before the anxiety escalates.

Enhancing Therapeutic Processes with VR/AR

Virtual Reality (VR) and Augmented Reality (AR) technologies offer powerful tools for enhancing therapeutic processes by creating immersive, controlled environments where patients can confront and manage their psychological challenges. These technologies can simulate real-world scenarios in a safe, controlled setting, allowing patients to practice coping strategies and receive immediate feedback on their physiological responses.

For example, a VR setup could simulate a social situation that triggers anxiety, such as speaking in front of a group. The patient can practice managing their anxiety in this virtual environment, with their physiological responses being monitored in real-time by wearables. The data collected can then be used to adjust the difficulty of the scenario, gradually exposing the patient to more challenging situations as their coping skills improve.

AR can also be used to overlay therapeutic exercises onto the real world, providing patients with interactive tools to manage their symptoms in their everyday environment. For instance, an AR app might guide a patient through a breathing exercise when it detects signs of stress, helping them to regulate their physiological responses in the moment.

The reinterpretation of transference and self-reflection through the lens of active inference has significant implications for therapeutic practice. Rather than focusing solely on uncovering unconscious conflicts or developing cognitive insight, therapists can use prediction errors as entry points for facilitating change. By gently confronting patients with discrepancies between their expectations and reality, therapists can guide them through the process of recalibrating their internal models. This aligns with the broader goal of reducing free energy, or the uncertainty and distress that arise from mismatches between expectations and experience.

Moreover, the integration of technology, such as wearable biofeedback devices and Virtual Reality (VR), offers new possibilities for enhancing this process. By providing real-time data on physiological states, therapists can gain insight into how patients' bodies are responding to emotional or cognitive discrepancies. VR environments, for instance, can simulate relational dynamics, allowing patients to experience transference in a controlled and immersive setting, thus enabling more precise interventions aimed at recalibrating interpersonal expectations.

These technologies not only enhance the therapeutic process but also provide valuable data that can be used to refine computational models of the patient’s condition, leading to more effective and personalized treatment plans.

?

Journaling and Well-being in the Context of Active Inference

Journaling, when viewed through the lens of active inference and the free energy principle, can be an effective tool for enhancing well-being by facilitating emotional regulation, belief updating, and reducing prediction error. Active inference posits that the brain continuously generates predictions about the world and updates these predictions based on sensory feedback, striving to minimize prediction errors—discrepancies between expectations and actual outcomes. Journaling plays a crucial role in this process by providing a structured means for individuals to reflect on their thoughts, emotions, and experiences, helping to reconcile these discrepancies.

Through journaling, individuals can actively engage in belief updating, a core mechanism of active inference. By writing about events and their corresponding emotional responses, people externalize their internal models of the world. This reflective process allows them to reassess and revise maladaptive beliefs or cognitive distortions, leading to more accurate predictions and better emotional regulation. For example, a person might journal about a negative social interaction, recognize cognitive distortions such as overgeneralization ("no one likes me"), and gradually replace these beliefs with more balanced predictions based on reality.

Moreover, journaling can assist in managing allostatic load, the physiological wear and tear caused by chronic stress. Writing about stressful experiences can help reduce the emotional intensity associated with stressors, leading to better emotional processing and physiological regulation. As individuals recount and process these experiences through journaling, they engage in a form of emotional regulation, lowering their physiological stress response and thus mitigating the impact of allostatic load on the body. This reduction in stress can enhance overall well-being by promoting a sense of psychological and physiological balance.

Additionally, journaling fosters self-reflection, a mechanism closely aligned with active inference, where individuals explore the discrepancies between their internal beliefs and the external world. This self-reflection not only helps to reduce prediction error but also enhances self-awareness and personal growth, promoting emotional resilience and well-being. By reflecting on both positive and negative experiences, individuals can recalibrate their expectations, leading to greater psychological flexibility and improved coping strategies in the face of future challenges.

In summary, journaling, through its capacity to facilitate belief updating, reduce allostatic load, and encourage self-reflection, serves as a powerful tool for enhancing well-being within the active inference framework. It enables individuals to engage in the continuous process of refining their predictive models of the world, promoting emotional regulation, resilience, and overall mental health.

?

Conclusion

Active inference, underpinned by the Free Energy Principle, offers a comprehensive framework that unifies psychological and neuroscientific perspectives. By viewing mental and bodily processes as predictive mechanisms aimed at minimizing uncertainty, we gain a more integrated understanding of cognition, emotion, and behavior. This approach not only bridges the gap between mind and body but also opens new avenues for more holistic and effective therapeutic interventions.

The paradigm shift from static to dynamic models of the mind and body holds the potential to revolutionize both theoretical and clinical practices in psychology. Instead of seeing the mind as a collection of separate parts, this approach views it as a continuously running system, always working to keep itself in balance. This perspective helps us understand how different aspects of the mind and body—thoughts, emotions, beliefs, and the self—are all interconnected and constantly influencing each other.

As we continue to explore and refine these ideas, the active inference framework promises to be a powerful tool for reimagining psychological concepts and enhancing our ability to promote mental well-being. By embracing this dynamic, process-oriented view of the mind and body, we can develop more effective strategies for understanding and treating mental health issues, ultimately improving the lives of those who struggle with these challenges.

?

?

?

References

1.??????????? Albarracin M, Bouchard-Joly G, Sheikhbahaee Z, Miller M, Pitliya RJ, Poirier P. Feeling our place in the world: an active inference account of self-esteem. Neuroscience of Consciousness. 2024;2024(1):niae007.

2.??????????? Biddell H, Solms M, Slagter H, Laukkonen R. Arousal coherence, uncertainty, and well-being: an active inference account. Neuroscience of consciousness. 2024;2024(1):niae011.

3.??????????? Chamberlin DE. The active inference model of coherence therapy. Frontiers in human neuroscience. 2023;16:955558.

4.??????????? Friston K. Computational psychiatry: from synapses to sentience. Molecular psychiatry. 2023;28(1):256-68.

5.??????????? Holmes J. Friston's free energy principle: new life for psychoanalysis? BJPsych bulletin. 2022;46(3):164-8.

6.??????????? Holmes J. Friston, Free Energy, and Psychoanalytic Psychotherapy. Entropy. 2024;26(4):343.

7.??????????? Holmes J, Nolte T. “Surprise” and the Bayesian brain: implications for psychotherapy theory and practice. Frontiers in Psychology. 2019;10:592.

8.??????????? Hopkins J. Free energy and virtual reality in neuroscience and psychoanalysis: a complexity theory of dreaming and mental disorder. Frontiers in Psychology. 2016;7:922.

9.??????????? Kirchner L, Eckert A-L, Berg M, Endres D, Straube B, Rief W. An active inference approach to interpersonal differences in depression. New Ideas in Psychology. 2024;74:101092.

10.???????? Limanowski J, Adams RA, Kilner J, Parr T. The many roles of precision in action. Entropy. 2024;26(9):790.

11.???????? Limanowski J, Blankenburg F. Minimal self-models and the free energy principle. Frontiers in human neuroscience. 2013;7:547.

12.???????? Lohr C, Hauke G. Piloting the update: The use of therapeutic relationship for change. a free energy account. Frontiers in Psychology. 2022:1306.

13.???????? Mende MA, Schmidt H. Psychotherapy in the framework of embodied cognition—does interpersonal synchrony influence therapy success? Frontiers in Psychiatry. 2021;12:562490.

14.???????? Miller M, Albarracin M, Pitliya RJ, Kiefer A, Mago J, Gorman C, et al. Resilience and active inference. Frontiers in Psychology. 2022;13:1059117.

15.???????? Parvizi-Wayne D, Sandved-Smith L, Pitliya RJ, Limanowski J, Tufft MR, Friston KJ. Forgetting ourselves in flow: An active inference account of flow states and how we experience ourselves within them. Frontiers in Psychology. 2024;15:1354719.

16.???????? Pezzulo G, Parr T, Friston K. Active inference as a theory of sentient behavior. Biological Psychology. 2024:108741.

17.???????? Pio-Lopez L, Kuchling F, Tung A, Pezzulo G, Levin M. Active inference, morphogenesis, and computational psychiatry. Frontiers in computational neuroscience. 2022;16:988977.

18.???????? Rabeyron T. Psychoanalytic psychotherapies and the free energy principle. Frontiers in Human Neuroscience. 2022;16:929940.

19.???????? Safron A, Sheikhbahaee Z, editors. Towards understanding persons and their personalities with cybernetic big 5 theory and the free energy principle and active inference (FEP-AI) framework. International Workshop on Active Inference; 2023: Springer.

20.???????? Smith R, Moutoussis M, Bilek E. Simulating the computational mechanisms of cognitive and behavioral psychotherapeutic interventions: Insights from active inference. Scientific Reports. 2021;11(1):10128.

21.???????? Smith R, Varshney LR, Nagayama S, Kazama M, Kitagawa T, Ishikawa Y. A computational neuroscience perspective on subjective wellbeing within the active inference framework. International Journal of Wellbeing. 2022;12(4).

22.???????? Solms M. The hard problem of consciousness and the free energy principle. Frontiers in Psychology. 2019;9:2714.

23.???????? Solms M, Friston K. How and why consciousness arises: some considerations from physics and physiology. Journal of Consciousness Studies. 2018;25(5-6):202-38.

24.???????? Waugh CE, Sali AW. Resilience as the ability to maintain well-being: An allostatic active inference model. Journal of Intelligence. 2023;11(8):158.

?