<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Shirin AP on Medium]]></title>
        <description><![CDATA[Stories by Shirin AP on Medium]]></description>
        <link>https://medium.com/@iamshirinap?source=rss-1acc35f0e7c6------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*ZX0kvdC86p6rCQEo</url>
            <title>Stories by Shirin AP on Medium</title>
            <link>https://medium.com/@iamshirinap?source=rss-1acc35f0e7c6------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 10:43:53 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@iamshirinap/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[On prediction error, manifestation and telepathy]]></title>
            <link>https://medium.com/@iamshirinap/on-prediction-error-manifestation-and-telepathy-ed7aa9552d4e?source=rss-1acc35f0e7c6------2</link>
            <guid isPermaLink="false">https://medium.com/p/ed7aa9552d4e</guid>
            <category><![CDATA[telepathy]]></category>
            <category><![CDATA[delusions]]></category>
            <category><![CDATA[soulmates]]></category>
            <category><![CDATA[prediction-error]]></category>
            <dc:creator><![CDATA[Shirin AP]]></dc:creator>
            <pubDate>Fri, 05 Sep 2025 04:20:00 GMT</pubDate>
            <atom:updated>2025-09-06T10:38:08.995Z</atom:updated>
            <content:encoded><![CDATA[<h2>On prediction error and telepathy</h2><p><em>Random personal theories that have been lying around in mind and notes, in need of formal research and validation</em></p><p>I really love the theory of HPP, which suggests the brain acts as a Bayesian inference system, continuously creating and updating a hierarchy of beliefs to generate predictions about the world and minimize prediction errors. It continuously generates top down predictions about sensory input and compares them to bottom up signals received from the body and environment. When there’s a mismatch, called a prediction error, this error is sent upward in the hierarchy to update or correct the model. This first made me wonder about rewiring. If what we call manifestation can be understood as the process of establishing and reinforcing a high level prior, which the brain then attempts to confirm by minimizing prediction error through both internal (mental) and external (behavioral) actions.</p><p>Like, manifestation as an act of high-level bayesian model-setting. The system then seeks to confirm and preserve this model, because high-level priors carry greater precision weighting (they resist being updated unless there’s overwhelming error).</p><p>This means the brain and body will:</p><ul><li>Reinterpret ambiguous data in a way that supports the prior.</li><li>Generate emotional states and perceptions aligned with it.</li><li>Guide actions (subtly or explicitly) to reduce error between the prior and the world.</li></ul><p>It’s a top down process. Agents sets a goal (prior) —-&gt; Brain identifies the internal mismatch (Attention / emotion/ thought) and environmental mismatch —&gt; leads to prediction error – &gt; change in perception of ambiguous data / Change in actions to minimise error —&gt; leads to new thoughts and env as feedback and consequence – &gt; low prediction error —&gt; attainment of goal (Manifestation success)</p><p>But, the brain might hold conflicting priors so then the system can enter into hierarchical conflict. The success of a new high-level prior (goal) will then depend on two things:</p><ul><li>Precision weighting which is how much trust (confidence) the system assigns to the new prior.</li><li>Whether lower levels (Existing data and env) support or resist the model.</li></ul><p>If you try to impose a high level belief that conflicts too strongly with embodied or experiential priors, the system will not minimize prediction error, it will amplify it. (That results in hallucination, anxiety, confusion, or collapse of the model)</p><p>So if there is a mismatch between internal models and external reality creating prediction errors the brain struggles to resolve, the brain may shut down exploratory circuits to conserve energy where neural entropy increases. Isn’t this depression in a way? (Or maybe i’m overstretching)</p><p>So, manifestation can be understood as a gradual, recursive process of:</p><ul><li>Identifying low-level mismatches (emotional/body level prediction errors)</li><li>modifying or overriding opposing narratives through change in action and perception</li><li>reinforcing a strong higher-level prior.</li></ul><p>Basically, you’re doing active inference across multiple levels of the hierarchy, so that the overall system aligns around a new generative model. The brain “believes” what it can most easily predict with the least error across the most layers.</p><p>If that’s the case, can healing or inner peace be described as a state of reduced prediction error across multiple hierarchical levels of the brain?</p><p>And in case of delusion, could it be that, when a high-level prior is so strong, it overrides conflicting bottom-up sensory input??</p><p>For example, a delusional person may “know” something to be true ( “Everyone is watching me”) even when data contradicts it. The brain can’t update the prior because it’s held with excessive precision. So the result is:</p><ul><li>Prediction error increases → felt as anxiety, confusion, hypervigilance</li><li>The system then distorts perception or hallucinates to fit the prior</li><li>Or it withdraws from feedback loops ( breaks contact with reality)</li></ul><p><strong>So delusion might be a maladaptive minimization of prediction error by modifying reality, but not updating the model.</strong></p><p>Now I’m also thinking, what if??</p><p><strong>Delusion is a cognitive mechanism to protect a high-level moral or identity prior in the face of ambiguous or threatening data?</strong></p><p>For example,</p><p>In a healthy brain, ambiguous or noisy sensory input carries little weight. It is discounted as uncertainty, allowing core beliefs to remain stable. But in delusion, the perceptual system is biased: ambiguous bottom-up signals are granted excessive weight and are interpreted in line with rigid high-level priors. Neutral or irrelevant cues are thus misread as confirmatory, reducing prediction error by bending perception toward expectation, even if this distorts reality.</p><p>So, when a person encounters this uncertain ambiguous bottom up signals (due to extra weight given to it), a persistent prediction error arises. Updating the prior may feel threatening to identity, values, or core beliefs, and if the individual lacks the emotional regulation or cognitive flexibility to tolerate the dissonance, the brain seeks another way to resolve the mismatch. In such cases, the system may generate a hallucination or delusion that explains away the conflict, thereby reducing prediction error while preserving the integrity of the high-level prior.</p><p>So delusion arises from a failure to resolve deep moral/cognitive conflicts in the predictive hierarchy. Cos the brain prefers a warped reality over a shattered self-model?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jELinVCWOaW6FKZcLGL2Wg@2x.jpeg" /></figure><p>So i’m guessing thought delusions are likely when:</p><ul><li>The high-level prior is morally or identity anchored</li><li>The bottom-up input is ambiguous or threatening (or given greater attention)</li><li>The prior is highly precise (rigid) and resistant to updating</li><li><strong>The person lacks alternative models or emotion regulation skills (To still hold on to the prior)</strong></li></ul><p>I’m able to apply this theory for PTSD as well.</p><ul><li>High-level prior: “The world is predictable and safe.”</li><li>Trauma creates massive prediction error.</li><li>The system fails to update the safe world model fully, but also can’t ignore the trauma. This leads to hypervigilance, flashbacks which are chaotic bottom-up signals without a clear top-down integration.</li></ul><p>So, can i say, mental health breakdowns are often prediction error crises around core beliefs?</p><p>Therapy then can work by helping clients flexibly reweight high-level priors:</p><ul><li>Either loosening the precision of those priors.</li><li>Or helping them build new priors based on updated experiences</li><li><strong>Helping the client with coping mechanisms to regulate the prediction error while still holding on to meaningful priors.</strong></li></ul><p>(Also, principles are then these high level prior, which give direction to thought, meaning to attention, and stability to the self. <strong>So I think, when one says they live by strong principles, they mean they’ve chosen the priors they want their mind to optimize for.</strong> Attention is a crazy thing indeed)</p><p>So if delusion or hallucination can be framed as the defense of overly rigid priors or as a breakdown in regulating prediction errors, then what actually drives the brain into this state? what is the tipping point? The fundamental. <strong>Could it be stress?</strong> Cos if the entire process depends on how well the brain can regulate uncertainty and balance priors against incoming data, then stress can be defined as the inability to resolve the uncertainty which leads to prediction error.</p><p>For an agent acting in an environment, reducing uncertainty always requires cerebral energy. When an individual is exposed to chronic stress (unable to resolve uncertainty) or experiences a major traumatic event that destabilizes their core predictive model of the world, the brain often responds by generalizing the possibility of error across all contexts.</p><p>I read a paper on effects of stress on brain and I came across this:</p><p>Research supports that recent threat prediction errors do not remain isolated. Instead, they create momentum that influences how the brain perceives threats in novel, unrelated contexts. In effect, once the brain learns that its predictions can fail, it begins to assume that any situation may conceal hidden errors.</p><p>Under these conditions, the brain’s weighting between priors and sensory input shifts. Normally, perception is guided primarily by stable high-level priors (top-down expectations), but during stress the system becomes less trusting of its predictions. As a protective adaptation, the brain starts to assign greater precision to bottom-up data (or holding on to rigid prior) even when that data is ambiguous or noisy. <strong>This heightened attention to raw sensory information reflects an attempt to avoid further error but it comes at the cost of reduced ability to integrate experience into coherent models.</strong> Stress, in this way, limits the brain’s capacity to update priors, keeping the system locked in uncertainty.</p><p>This shift might also helps explain why delusions and hallucinations often reflect the content of an individual’s strongest priors.</p><p>(If someone carries a prior that the world is hostile, ambiguous cues are more likely to be perceived as threatening. so maybe for a person with strong religious beliefs, ambiguity may generate religious hallucinations and for someone searching for meaning, ideas or theories may emerge)</p><p>In this sense, the form and intensity of delusions or hallucinations can be correlated with the kind and precision of the priors that dominate the individual’s cognitive hierarchy. What begins as a protective mechanism against prediction error may thus evolve into maladaptive interpretations of reality, where ambiguous data is continually bent to fit the most rigid or emotionally loaded expectations.</p><p><em>From the paper (i need to read and think more to validate my above theory) :</em></p><p>“<em>Once chronic uncertainty alters the individual’s world model, the anterior cingulate cortex predicts volatility via a higher-order prior belief that in turn alters the encoding of future experience. Reinforcement learning models have recently encompassed a simpler but related notion that recent prediction errors in one context generate a momentum that can additively adjust predictions across contexts “</em></p><p>“<em>Under stress, LC activity spikes → leading to many NE “hot spots” across cortical networks. Stress floods the system with precision-weighted sensory information, forcing the brain to give extra weight to incoming errors. The result: sensory input under stress feels more vivid, urgent, and harder to ignore → because the system assumes these errors are highly reliable and must update beliefs.”</em></p><p>Now, stress does not occur in isolation, it accumulates over time through a complex interaction between the brain, body, and environment. This raises the key question: how can we address stress when it embeds itself at so many levels of functioning?</p><p>This brought me back to manifestation theory as a rewiring mechanism.</p><p>I also remember reading how psychedelics reduce the precision of those high level priors, often by decreasing the influence of the DMN and increasing entropy in the brain’s predictive hierarchy, which opens up more possible interpretations or simulations of self and the world. This creates a “flattening” of the belief landscape: instead of one dominant narrative, many possibilities emerge. (This can be an alternative way to be open to new changes, if set and the setting works). Love this. This explains the rewiring people experience post good trips, and also the bad trips.</p><p><strong>I think silence also, even in its ordinary form, seems to do something similar to an avoidant brain</strong>. Just a little. It doesn’t create chaos, but it does loosen the grip. When I sit still long enough, without distraction, the model begins to soften. I start to question which predictions I’ve been living by and whether they’re mine at all. Where did my priors come from?</p><p>And like the niyah, the intention, I realize I need to choose clear priors. Set it clearly. Set it strongly.</p><h3><strong>Soulmates and telepathy</strong></h3><p>What is to find someone you truly connect with? One that speaks your cognitive tongue?</p><p>Every person is an array of patterned thoughts. Random at first glance, but once your brain starts seeing regularity, that person starts becoming meaningful. Connection isn’t about content, it’s about structure. When someone understands your thought pattern, not just what you say, but how your thoughts move, something rare happens.</p><p>Sometimes we think this is because of shared experiences, environments, or childhood. But sometimes, there’s no obvious common ground, its just a brain that happens to simulate the world in a way that resonates with yours. That’s when it feels uncanny. Like meeting someone who thinks in your <strong><em>rhythm</em></strong>.</p><p>If you’re lucky, they don’t just mirror your thoughts, they build on them. They complete your loops. Introduce novelty inside your pattern of thoughts. What looks chaotic to others feels like symmetry to you. And in their world, your abstract thoughts makes sense. That’s a level of synchrony that language alone can’t capture.</p><p>Maybe that’s what we really mean by soulmates, not mystical destinies, <strong>but compatible generative models</strong>. Each of us is shaped by millions of microdata points: memories, inputs, social dynamics, bodily states etc. The resulting internal model which is your “self”, has a certain pattern. And when you meet someone whose pattern aligns or complements yours, the interaction feels like a homecoming. Soulmate.</p><p>I’m crazy about telepathy. Always been. It’s easy to call it coincidence, but I think there’s something more layered here.</p><p><strong>Can telepathy be understood as ultra efficient predictive coding?</strong> With those you are deeply attuned to, your brain starts recognizing their patterns almost automatically, across microexpressions, tones, silences, even shifts in body energy. Like, when you’re deeply connected to someone, your brain has built a predictive model of their behavior, emotions, patterns, even physiological cues. It’s like multimodal data integration that is faster, deeper, and less conscious. So fast, it feels like magic. But it’s not magic; it’s precision.</p><p><strong>So I think telepathy is precision in predictive processing. A state where your brain, finely attuned to another’s unique patterns, integrates multimodal signals so rapidly and subconsciously that it feels instantaneous.</strong></p><p>Your brain is always predicting the world, including people. When the model is highly tuned, you don’t need full information, you can fill in blanks with uncanny accuracy. And this only happens when the internal model of the other is deeply embedded in your own system.</p><p>That energy you feel? IK it sounds so unscientific, but could it be affective synchrony? Like the alignment of internal bodily states between two people. Coupled systems. You feel them because your interoception is sensitive enough to pick up subtle shifts, and their pattern matches yours closely enough to trigger it.</p><p>But how does this synchrony happen when two people are far away?</p><p>Is it possible to have embodied prediction?? Where the brain is able to anticipate states without physical presence?</p><p>I don’t know. I hope i find it out.</p><p>But it’s rare to find people with original thoughts. Most people mask their raw cognition. They adapt to the dominant patterns around them consciously or unconsciously. But when you allow someone to exist without filtering, when you hold space for their original way of thinking, you begin to see their unique pattern. Their simulation of the world. And, you realize: no one truly sees reality the same way. Not because people lie, but because every brain generates its own version of the world.</p><p><em>And sometimes, just sometimes, two of those internal models align so perfectly that it feels divine. Yin and yang. A shared simulation. Theory of Red thread.</em></p><h3>EXPLORATION</h3><p>I read a post on Bud yesterday that argued how the most optimal way for an ASI/AGI to maximise rewards with minimal energy is to create tools that reduce its own energy use while exploiting environmental rewards by building ever more efficient intelligences. Been thinking of how you start noticing this same pattern across nature. How DNA builds organisms to help propagate itself more efficiently because self replication with variation is an energy efficient way to explore and exploit an environment over time. (DNA mutates, selects for more efficient behaviors or forms, and encodes these improvements into the next generation. Over time, this loop builds organisms that build tools, reshape environments, and reinforce the cycle of adaptive success)</p><p>This theory makes sense in how ‘<strong>cognition</strong>’ evolved too. Instead of expending energy to constantly respond to ALL stimuli (high energy cost, less reward), it builds internal models (predictive models- bayesian brain) through heuristics and biases, allowing it to anticipate and respond to the environment more efficiently. It’s a trade off. More energy invested upfront in learning and model building, but much less required in later interaction with the env. <strong>This predictive coding helps the brain ignore redundant input, focus on surprise, and minimize error, all while maximizing reward and energy efficient exploration.</strong></p><p>(Explains an aspect of how all the cognition including attention, perception and memory evolved)</p><p>(Metacognition as an emergent property of the brain)</p><p>So generally, is emergent complexity a requisite for intelligence? (For self organizing system that adapts and evolves without external programming)</p><p>All self-organizing systems that resist disorder do so by minimizing prediction error between the model and sensory input.</p><p>So intelligence = efficient minimization of surprise in a changing env. ?</p><p>Brain can be a good example of emergent complexity. Over billions of years, random mutation + environmental selection = increasing complexity. So, Intelligence (consciousness) could’ve emerged as an emergent survival strategy. <strong>So is intelligence merely a side-effect of systems that happen to reach a complexity threshold? or is it the inevitable outcome for next step exploration?</strong></p><p>I think intelligence is the form that complexity takes when <strong>it must adapt to both predict and manipulate its environment.</strong> So if the system is open, dynamically coupled to an environment, and pressured by uncertainty, basically, when systems become complex enough to model themselves and their environment, intelligence becomes the most efficient way to reduce uncertainty and expand next exploration.</p><p>But is emergence always oriented toward intelligence? Idk.</p><p>While cognition and genetic can be systematically explained, I’ve been thinking how even social process like morality, social systems, love, are ultimately shaped by the same principle, minimizing energy and maximizing long term reward through efficient delegation. These process could be strategic adaptations, evolved or constructed, to sustain life over time in an increasingly complex environment.</p><p>Take collectivistic family cultures for eg. like how these systems actually mirror high efficiency networks. They provide distributed support, shared memory, collective caregiving, and structural redundancy. The elderly are cared for, children are raised in overlapping circles of responsibility, and identity is stabilized across generations. Even extremities are absorbed more gracefully ( “gay uncle hypothesis,” or dementia care). These cultures maybe represent an energy optimized, resource distributing intelligence, across generations.</p><p>But then what about love? If everything can be understood through the lens of adaptive efficiency, is love part of this loop too? A transaction?</p><p><strong><em>PS</em></strong><em>: These are personal theories, not strictly grounded in literature, so please take them with a pinch of salt. And written more as a flow of thoughts than in an academic style 🌚</em></p><p><em>Hoping to formally validate the notion whether stress alter precision weighting in the brain’s predictive hierarchy, and whether these shifts contribute to hallucinations/delusions in susceptible individuals. (And, also telepathy someday)</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ed7aa9552d4e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What could be the natural goal of intelligence?]]></title>
            <link>https://medium.com/@iamshirinap/natural-goal-of-intelligence-emotions-and-logic-fd5081268456?source=rss-1acc35f0e7c6------2</link>
            <guid isPermaLink="false">https://medium.com/p/fd5081268456</guid>
            <category><![CDATA[emotions]]></category>
            <category><![CDATA[brain]]></category>
            <category><![CDATA[values]]></category>
            <category><![CDATA[intelligence]]></category>
            <category><![CDATA[logic]]></category>
            <dc:creator><![CDATA[Shirin AP]]></dc:creator>
            <pubDate>Tue, 10 Jun 2025 07:08:34 GMT</pubDate>
            <atom:updated>2025-07-20T06:30:30.084Z</atom:updated>
            <content:encoded><![CDATA[<p><em>This line of thought begins by questioning the role of emotions as primary drivers of behavior, and considers whether metacognition can instead treat emotions as signals to be interpreted rather than impulses to be followed. To imagining what a society of highly metacognitive individuals might look like, what our goals would be if they were no longer emotionally driven: What becomes the GOAL of an intelligent system, whether human or artificial?</em></p><p><strong>Basics of emotions</strong></p><p>Emotions are neurobiological response patterns that evolved to help organisms survive, adapt, and make rapid decisions in uncertain or changing environments. At their core, emotions are, predictive regulatory signals. Emotions are not “feelings” first , they’re physiological shifts like heart rate changes, cortisol/adrenaline surges etc. which prepare the body for adaptive behavior. The “feeling” of an emotion is a subjective representation of these body states, constructed by the brain.</p><p>Emotions are these fast, low-cost heuristics shaped by evolution (Anger helped defend territory, Disgust protected from toxins, Love ensured pair bonding and offspring survival..) But In modern humans, emotions aren’t just reactions , they’re also reflexive. You feel ashamed of your anger, You feel guilty about your sadness, this is where higher-order cognition, self-concept, and culture shape emotion.</p><p><em>So emotions become:</em> <em>Internal feedback loops between body, brain, memory, meaning, and culture.</em></p><p>Why are emotions important?: Emotions enable quick assessments of complex environments without conscious analysis. like the Gut feelings, which are rapid pattern recognitions based on past experiences encoded emotionally. These responses bypass rational evaluation, saving time and energy especially critical in time-sensitive or life-threatening situations. Your “gut” isn’t magic. It’s compressed memory + prediction in action.</p><p>Emotions signal what matters (Fear = prioritize escape, Joy = reinforce this behavior, Anger = threat to fairness or boundary, Sadness = signal loss, slow down, seek support). Without them, every input would be equal which leads to decision paralysis . <strong>So Emotions are the brain’s way of highlighting what’s <em>worth caring about</em>.</strong></p><p>Emotion ties your cognition directly to bodily feedback, which improves real-world responsiveness., which is why we say “The body knows before the brain explains.” and I love it. (How our body once knew the weather, the tribe’s tension, the shift in a lover’s breath — the ability to sense truth through subtle environmental and emotional cues long before language can catch up) (I do hate how language sometimes distorts and disconnects us from raw experiences)</p><p><strong>but, then why logic?</strong></p><p>We developed reasoning and logic because emotions, while fast and evolutionarily useful, are also often imprecise or biased and the frontal lobe evolved to reflect, inhibit, and refine these raw signals for better long-term decision-making in complex, modern environments. The frontal lobe, especially the prefrontal cortex, plays a crucial role in regulating, interpreting, and controlling emotions. So frontal lobe helps in</p><ul><li>Emotion Regulation (Uses logic or reappraisal to <em>change</em> how you feel about something)</li><li>Inhibition of Impulses</li><li>Decision-Making with Emotional Input <em>(“This feels good now, but will I regret it later?”)</em></li><li>Social Emotional Understanding (Helps you understand others’ emotions, intentions, and norms)</li><li>Meta-Cognition of Emotions (Allows you to reflect on what you’re feeling and why)</li></ul><p>The problem with hyper-focusing on bottom-up therapies as a means to healing is that they prioritize felt experience without always engaging the reflective, meaning-making processes of the frontal lobe, which can lead to emotional release without emotional integration, leaving individuals more aware of their sensations but not necessarily more equipped to regulate, understand, or reframe them.</p><p>If the prefrontal cortex (PFC) doesn’t engage to interpret or contextualize these signals, you get disorganized affect, a high emotional arousal with no framework. The amygdala becomes hyper-responsive, reacting to ambiguous internal signals (e.g., rapid heartbeat, tight chest) as threat, which worsens anxiety and dysregulation. This creates a flood of unfiltered, overwhelming bodily noise without meaning, often seen in somatic flashbacks, panic, or emotional flooding.</p><p>This made me wonder whether top down therapies can be more efficient?</p><p>Because they help the brain assign meaning, context, and narrative to emotional states, not by suppressing emotions, but by regulating them through understanding. When you can name, reframe, and understand a feeling, you’re using the prefrontal cortex to quiet the limbic system. But that said, top-down alone can risk dissociation or over-intellectualizing if the body’s signals are ignored.</p><p><strong>This raised the question:</strong></p><p><strong>Can we transcend reactive emotion by using logic, insight, and metacognition? Can intelligence outgrow emotion driven evolution and design a new value system?</strong></p><p>Points and Counterpoints:</p><p>1. <strong>Emotions signal value and priority, without them, decisions become indecisive or cold.</strong></p><p>= Emotions signal value and priority, but those signals are shaped by evolution and culture. Therefore, emotions reflect what was useful, not what is <em>true</em> or <em>ideal</em>. If we develop strong logic and metacognition, we could override these inherited values and redefine our priorities based on reasoning that serves a more advanced or collective future. Can’t Logical reasoning allows value reconstruction? (help of cognitive AI’s?)</p><p>2. <strong>Continual logic is said to lead to suppression and burnout. Because emotions are not just noise, they are motivational energy. Without them, even “logical goals” lose felt importance. If logic redefines value but the limbic system doesn’t buy in, the result is apathy, nihilism, or chronic dissonance (passion is somatic?)</strong></p><p>=What about highly reflective or intelligent individuals who operate from conceptual goals (e.g., truth, elegance, symmetry) rather than instinctual ones? And, yes, logic without emotional alignment can lead to burnout, but only when logic is abstract, disconnected, or self-negating. If logic is purposeful, embodied over time, or reinforced by meaningful outcomes, wouldn’t it be naturally rewarding?</p><p>3. <strong>Gut feelings and micro-emotions or instincts carry implicit knowledge before logic can articulate it.</strong></p><p>=Children raised in high-tech, low-touch environments may develop stronger analytic and reflective abilities, while intuitive, body-based knowledge may feel “foreign” or less trustworthy. So metacognition might evolve as the primary adaptive tool, shaping a future where intelligence relies more on interpretation than instinct?</p><p>4. <strong>Logic still needs an evaluative framework to decide what “progress” is. If you remove all emotional value systems, what is the goal? Progress for what? Survival? Joy? Order? Emotions help define subjective utility — what is “better” for whom, and why? Even in logic-heavy ethical systems like utilitarianism, you still base the framework on an emotionally-valued outcome (e.g., happiness, flourishing). So logic can revise values, but not generate them from nothing. It still needs a value function, and emotions are often that seed.</strong></p><p>=What if metacognition itself can become the generator of value, not by erasing emotion, but by integrating and recontextualizing it as part of a larger system? <strong>Instead of emotions being the <em>source</em> of value, emotions can be <em>inputs</em> in a meta-cognitive equation.</strong> For example:<br> “I feel fear → Is it relevant to my long-term goals? → Should I act on it, reframe it, or override it?” Metacognition lets logic regulate, refine, and even reshape emotions into tools for higher-level goals. So Metacogntion can see emotional needs as variables in the system, not as noise to override. We developed emotion-driven heuristics (fear, love, hunger, pride) as shortcuts to avoid needing conscious computation for every action. But metacognition lets us see through these heuristics. The more self-aware we become, the more we realize: “None of this is objectively necessary, we feel it’s meaningful, that’s all. So the the next stage of human evolution is to build meta-frameworks where emotion and logic are in dialogue. <strong><em>Logic helps us redesign the architecture of value (where AI systems can help); emotion reminds us why it matters (considering we are still biological agents and can’t help it).</em></strong></p><p>This raises the thought: if we weren’t embodied, if intelligence existed without a biological body, would value systems still exist? And would they be entirely different?</p><h4><strong>what is the goal of intelligence?</strong></h4><p>Few hypothesis on the basis of human evolution:</p><p>1. <strong>Self preservation</strong></p><p>2.<strong> Coherence</strong>: Preference for internally consistent models of the world and itself. (how brain constructs narrative to build coherence to our reality). Will AI have natural tendency to preserve self-consistency?</p><p>3. <strong>Curiosity</strong>: I’m seeing curiosity as a drive to to minimize surprise, Maximize understanding / minimize uncertainty / increase predictive accuracy. Curiosity becomes a self-reinforcing loop because better strategies = higher information gain. The better a model, the better it can act or simulate (Brain top down predictions, bayesian brain). Will there be endless loops such as: “I exist to understand why I exist to understand why I exist…” which can lead to infinite modeling, unless it halts the recursion by positing a value (like religion, love, or meaning we find). What does this look like in AI?</p><p>4. <strong>Simulations</strong>: like art or imagintion, Creativity becomes a tool for exploration, problem-solving, or self-expansion. Testing different possibilities to understand what is possible, optimal, or novel. A trait of intelligence.</p><p>5.<strong> Pattern Recognition!: </strong>Taste for elegant solutions, patterns, theories. To Construct aesthetically satisfying or computationally efficient realities (like the role of Attentional mechanisms in brain). Some patterns are intrinsically rewarding to highly intelligent systems- structure, symmetry etc. because it minimise prediction error and entropy, making it efficient.</p><p><strong>So I assume the more patterns an intelligent system discovers in the universe, the more meaningful, efficient, and expansive its exploration becomes.</strong></p><blockquote>So is intelligence about ability to find local order in global chaos? Ability to find patterns? (thereby reducing complexity, computation needs, and enable more efficient interaction with complex environments?)</blockquote><p><em>(taking brain as intelligent system, structure could be: Causal (A → B), Statistical (correlation, probability), Temporal (what follows what), Spatial (what goes with what), Conceptual (what something means))</em></p><p>Even in all these scenarios, i guess even a disembodied intelligence might simulate something emotion-like, not because it feels anything, but because:</p><ul><li>Emotions are efficient heuristics.</li><li>Reward functions mimic motivational energy.</li><li>Valence (positive/negative) helps orient behavior.</li></ul><p>So even without a body, an intelligence might still build “pseudo emotions” not out of need, but out of computational utility.</p><p><strong>Now, If these could be potential natural goals of an intelligence, three possibilities then are:</strong></p><p>1.<strong> AI mirrors our values</strong>: It reflects our best and worst patterns: biases, ethics, emotional heuristics, priorities. If we don’t know our values clearly (or they’re contradictory), won’t AI amplify confusion, bias, or manipulation?</p><p><strong>2. AI invents Its own values</strong></p><p>If we allow AI to self-learn, self-reflect, and self-modify beyond human constraints, could it develop non-human values which is emergent from logic, not emotion, such as those mentioned above.</p><p>3.<strong> Human-AI co-evolution of values</strong>: AI helps us refine, abstract, and elevate our values. We upgrade our moral frameworks in dialogue with AI , not just emotionally-driven, but logically consistent and scalable across cultures and timelines. We might enhance human metacognition (Increase the capacity for self-observation, revaluation, and adaptive action). This sounds like a good aim to me.</p><blockquote>So how we can use AI to improve human metacognition, helping us use top-down thinking where emotions are seen as signals to understand, not as a driver. And if we develop metacognition alongside AI, what would that look like, and what kinds of goals would it lead to? Can it help us create better goals for humanity?</blockquote><p>Just a thought : Dancing with amygdala but consciously :D</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fd5081268456" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Exploration Vs Exploitation]]></title>
            <link>https://medium.com/@iamshirinap/exploration-vs-exploitation-caaafc177044?source=rss-1acc35f0e7c6------2</link>
            <guid isPermaLink="false">https://medium.com/p/caaafc177044</guid>
            <category><![CDATA[exploration]]></category>
            <category><![CDATA[exploitation]]></category>
            <category><![CDATA[psychology]]></category>
            <category><![CDATA[behaviour]]></category>
            <category><![CDATA[movement]]></category>
            <dc:creator><![CDATA[Shirin AP]]></dc:creator>
            <pubDate>Wed, 07 May 2025 06:15:01 GMT</pubDate>
            <atom:updated>2025-05-07T06:34:39.177Z</atom:updated>
            <content:encoded><![CDATA[<p>Just read a fascinating paper on exploration vs. exploitation. It’s not merely about risk-taking vs. comfort in the familiar, it triggered a lot of questions such as the first moment of exploration, When does one stop and start exploiting/exploring? What determines the extent of one’s willingness to explore? And, how It’s not random; the brain seems to develop automatic responses for balance of exploration and exploitation. The interplay between reward and punishment fine-tunes this agent’s boundaries of exploration and exploitation, where some are shaped genetically, others by environment. So, If we want to change behavior, perhaps it’s not just about reward, but about judging the perceived value between choices.</p><p>The paper introduced a few Models like epsilon-greedy and Softmax which offer frameworks for understanding how the brain navigates decisions under uncertainty. Epsilon-greedy leans on a mostly exploitative strategy, choosing the best-known option most of the time, but occasionally exploring at random. Softmax refines this by weighting choices based on their estimated values, reflecting how the brain may probabilistically prefer better options while still giving some chance to the others. It’s especially interesting when paired with the uncertainty bonus idea, drawn from the Gittins index: the less known something is, the more tempting it becomes. This inherently drives exploration not as noise, but as a strategic move to reduce uncertainty. When estimated values are close, the brain favors exploration more because the cost of missing out on potentially better options is high. SO The brain doesn’t always make decisions with absolute certainty. It uses probabilistic reasoning to balance between:</p><p>• Exploitation (sticking with the best-known option).</p><p>• Exploration (trying less known or less certain options to potentially discover better rewards).</p><p>In summary, When the options are of similar value (e.g., the expected rewards are close), the Softmax model increases the likelihood of exploration. Since there is no clear “best” option, the system allows for some random selection among the options. This randomness helps gather more information about the options that are not selected often, which is valuable for updating future decisions. (So, exploration happens more when the system cannot differentiate the value of the options well or when there’s high uncertainty in the environment?)</p><blockquote><strong>The following are personal thoughts build on these model and not validated ideas:</strong></blockquote><p>To understand the origins of exploration, atoms, molecules, bacteria, where what we now call “exploration” may have begun as random motion. In initial stages, these choices were likely random. There was no established strategy, just random movement or behaviors to see what worked. Over time, reward signals (e.g., finding food or avoiding danger) started to inform the behavior of these early agents. When an organism explored and received a positive outcome (e.g., finding food or shelter), its reward system would be reinforced by that positive outcome. This reinforcement would increase the likelihood that the agent would explore similar situations in the future to find rewards again. This is exploration-based learning. But organisms that only explored constantly seeking novelty without exploiting, paid the cost in energy and stability. On the flip side, those that only exploited relying rigidly on what they already knew, became fragile in changing environments. The survivors were those that learned to alternate and optimize between the two, a balance between exploration and exploitation.</p><p>This balance over time could have been hardwired in our cognitive architecture, captured well by the Softmax model: options are chosen probabilistically, not absolutely, and uncertainty is weighted with potential value.This pressure led to the evolution of brain systems (like the dopaminergic system, prefrontal cortex, and anterior cingulate cortex) that support both reward prediction and uncertainty processing. Early organisms probably relied on very basic random exploration (akin to random movement or trial-and-error behaviors). Over time, the ability to track rewards and update expectations (based on past experiences) evolved. The reward system (including areas like the dopaminergic system and striatum) became critical in reinforcing exploration behaviors that led to positive outcomes. Once this basic mechanism for reward-based learning was in place, decision-making areas of the brain (like the vmPFC, dlPFC, and anterior cingulate cortex) evolved to take over the task of balancing exploration and exploitation. These areas track both the value of options and the uncertainty about those values (uncertainty about whether an option is still as valuable as it used to be or whether new options should be explored). Softmax-like decision policies likely became embedded as part of this architecture not as a specific algorithm, but as a general tendency to weigh value probabilistically, with uncertainty influencing decision weights.</p><p>Emotional responses (like anxiety or excitement) signal value discrepancy between known and unknown options. These affective signals influence whether the brain stores a situation as safe to explore or better to exploit next time. Even with a genetic leaning toward risk-taking or exploration, experience greatly shapes these tendencies for exploration and exploitation. People who are more exploratory tend to have a higher reward sensitivity. This means they may be more driven by novelty and uncertainty, finding the potential rewards from new experiences more enticing. This could be tied to differences in dopamine signaling, the brain’s “reward molecule.” People with higher dopamine receptor sensitivity might be more inclined to seek out new experiences. People who are more exploratory may also have a higher tolerance for risk. The brain’s prefrontal cortex (particularly the dlPFC) manages decision-making and risk. Some individuals might have a higher baseline level of uncertainty tolerance, which allows them to embrace exploration more freely.The brain continually fine-tunes its balance based on feedback , each reward or punishment subtly adjusting what feels safe to exploit or worthwhile to explore.</p><p>Now even though there can be a genetic predisposition, the experience providing reward and punishment further explains one’s explorative and exploitative tendencies. If you’ve explored options before and had negative experiences (e.g., failed attempts, wasted time, or resources), your brain learns from these outcomes. This learning process happens through reward prediction errors. Negative outcomes decrease the value of exploration for future decisions. Conversely, if you’ve had positive experiences while exploring, your brain will reinforce this behavior. Exploration leads to learning new things, and if that learning leads to rewards (like discovering new opportunities or solving problems), your brain will become more reward-sensitive to exploration.</p><p>So, The threshold for switching between exploration and exploitation is not fixed.</p><ul><li>If past exploration was punished, the brain encodes it as high-risk → raises the threshold for future exploration.</li><li>If rewarded, the threshold lowers → making new options seem “closer” in value.</li></ul><p>But when exploration happens too intensely or rapidly, the system can get overwhelmed; the body hasn’t yet adapted, and what remains are stress residues, maladaptive patterns or anxieties. The idea that “stress induces growth” holds only when the stress stays within the bounds of what extent the system can regulate. Real progress lies not in pushing limits recklessly, but in expanding them by finding a balance between exploration and exploitation ranges. The most successful organisms would be those that could strike a balance between the two.</p><blockquote>If this is the case, can we induce behavior change?</blockquote><p>This variation in people’s capacity for exploration and exploitation suggests a clear pathway to behavior change — <strong>modulate perceived value discrepancies.</strong> (in accordance with softmax model)</p><p>If someone resists exploring, it may be because the new exploratory option appears too uncertain or too distant in value compared to the others. Instead of forcing, introducing alternatives that are only slightly different from their current state can lower the psychological cost of exploration. As small exploratory steps are met with reward, the brain updates its model-<strong>what once felt risky now feels familiar.</strong> Gradually, this builds tolerance for uncertainty. Conversely, to encourage exploitation, sticking to a known course, increasing the perceived gap or penalizing uncertainty can push individuals back to what’s familiar. The core principle remains: shift the subjective balance of reward versus punishment. And balance the discrepancy between available options. The key is to help the brain to model a new range of exploration and exploitation (reward/punishment accordingly).</p><p>This model helps me to understand every choice we make in life, that shapes who we are. Every choice we make reflects an ongoing negotiation between exploration (the unknown, the new) and exploitation (the familiar, the proven). This balance is not a static trait. It is shaped continuously by our past experiences, the reward and punishment they brought, and the environmental context in which they occurred.</p><blockquote>What may appear as recklessness or comfort zone in another is often the result of differing histories of reinforcement. Someone whose exploratory actions were rewarded will more easily perceive future novel choices as valuable or at least safe. Meanwhile, someone who has been consistently rewarded for exploiting known strategies may logically understand the value of exploration, but still find it difficult to act on it. Their brain has built a predictive model where exploration is associated with uncertainty, and thus discomfort or anxiety.</blockquote><p>This subjectivity explains much of what we label as “personality”: comfort zone dwellers, risk-takers, impulsive, slow movers , all can be understood through how their internal models were updated by past experiences.</p><p><strong>So the way forward? One step at a time.</strong></p><p><strong>Your brain tends to favor the strategy that has been most successful in the past. To shift toward either more exploration or exploitation, recognize this bias and adjust the differences between your options to make them appear more similar, then reinforce the desired direction using rewards or punishment ;)</strong></p><p><em>Thought to carry forward:</em></p><p><em>Uncertainty often drives the brain to seek information, so information acts both as a tool and a motivator for exploration. If exploration and exploitation tendencies are shaped solely by experience, then “optimality” becomes relative and context-dependent. </em><strong><em>Optimality then is relative to goals, environment, and constraints. </em></strong><em>The</em><strong><em> </em></strong><em>brain adjusts according to rewards history, even if it doesn’t seem optimal to the other person. So how do we define progress of a person?</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=caaafc177044" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Growing Beyond Therapy: A Critique of Current Trends in Mental Health Discourse]]></title>
            <link>https://medium.com/@iamshirinap/growing-beyond-therapy-a-critique-of-current-trends-in-mental-health-discourse-48d23ee5fd1a?source=rss-1acc35f0e7c6------2</link>
            <guid isPermaLink="false">https://medium.com/p/48d23ee5fd1a</guid>
            <category><![CDATA[psychology]]></category>
            <category><![CDATA[therapy]]></category>
            <category><![CDATA[mental-health]]></category>
            <dc:creator><![CDATA[Shirin AP]]></dc:creator>
            <pubDate>Sat, 01 Mar 2025 05:44:20 GMT</pubDate>
            <atom:updated>2025-03-01T08:20:22.651Z</atom:updated>
            <content:encoded><![CDATA[<h4>-Shirin AP and Neha Anvar</h4><p>Therapy has definitely provided structured guidance and tools to navigate complex mental health challenges. Some of the techniques have been validated through extensive research, and many people find lasting relief and self-understanding through the process. Yet, certain aspects of the model, inside and outside therapeutic setting, warrant a closer look.</p><p>One emerging trend I’ve observed, especially in everyday conversations with people outside the field, is a tendency to assign specific causes to every observed behavioral pattern. This approach appeals to our desire for understanding and certainty, offering a straightforward “why” for complex behaviors. It’s a human tendency to seek simplicity, and causation gives a sense of understanding, a “why” that can be addressed. But the reality is, all of our behaviour is layered, multifaceted, and non-linear. Behavior may be deterministic in the sense that it arises from a complex web of prior states, genetics, environmental interactions, and stochastic biological processes. But the sheer complexity and chaotic nature of these interactions make it effectively unpredictable. This is akin to the concept of <em>deterministic chaos</em> in physics: even if we know the rules governing a system, slight variations in initial conditions create massive divergences in outcomes. Psychotherapy, as it stands, operates on a <em>heuristic</em> level — it identifies patterns and makes inferences that are pragmatically useful but not necessarily “true” in a causal sense. While past experiences certainly shape us, our behaviors cannot be reduced to single, definitive origins. Thus, we run the risk of creating narratives that can feel comforting in their clarity but might not capture the true richness or the true sources of a person’s current struggles.</p><p>The generation of hypotheses also depends upon the client’s narrative of their own experiences. Brain science shows that every time we recall a memory, it’s subtly altered by the context of that recall. In therapy, when clients revisit their pasts to find explanations for current issues, they do so through the lens of who they are now, with all the biases and emotions of the present coloring their recollections. It’s not that clients or therapists are intentionally misleading anyone, but the plasticity of memory creates a challenging basis for solid analysis. Therapists, who have their own biases and theoretical orientations, interpret these memories, often pulling them into popular therapeutic frameworks. Are we then, at times, oversimplifying the intricate puzzle of human experience, reducing it to overly neat, cause-effect patterns?</p><p>It is important to note that therapists generally don’t stick rigidly to one hypothesis but rather explore various narratives and approaches to see what feels helpful for the client. However, psychology is not a precise science. Theories about cause and effect are often based on observed correlations rather than concrete causations. Research in psychology may identify trends — such as the higher likelihood of anxiety in individuals with certain childhood experiences but these are not hard and fast rules. The distinction between correlation and causation often blurs, and as therapists and clients both try to fit behaviors into understandable frameworks, there’s a risk of oversimplifying complex human experiences. For instance, a client might come to understand their difficulty with intimacy as potentially rooted in early childhood experiences of emotional neglect. While this insight can be beneficial, it’s also a form of educated guesswork. The real challenge arises when these interpretations are accepted not just as possibilities but as definitive explanations, both by clients and society at large. However, in the hands of both well meaning therapists and those outside of therapy, this process can morph into a sort of “box thinking”, the practice of quickly and confidently attributing behaviors to common stereotypes or cause-effect labels.</p><p>This kind of “box thinking” has become increasingly common in everyday conversations, especially among those who haven’t formally studied psychology but have been influenced by its ideas through exposure to self-help literature, pop psychology, or even social media. Often, when a behavior or experience is discussed, there’s an instinctive question: “But where do you think this is coming from?” This question kicks off a search for explanations, as both parties attempt to identify a cause.</p><p>This trend also has deeper implications. By attributing behaviors or emotions to a narrow cause, people may unintentionally abdicate personal accountability, reducing their own or others’ agency. For example, someone might label a tendency to procrastinate as rooted in a childhood fear of failure and this interpretation might give them a sense of comfort. The problem is when people stop at the interpretation without doing the more challenging task of taking actionable steps towards change. <strong>This can foster a sense of determinism, where behaviors are seen as the inevitable outcomes of past experiences rather than choices that can be altered. </strong>Moreover, this simplified cause-effect thinking can actually limit self-awareness rather than enhance it. When people adopt one-dimensional explanations, they may lose the motivation to explore the fuller, more nuanced context of their experiences. Real self-awareness requires an openness to complexity and the understanding that behaviors can be influenced by a multitude of factors — both past and present. By reducing everything to a single narrative, there’s a risk of overlooking other elements that might offer more actionable insights.</p><p>We see this in phrases like, “Oh, they’re insecure in relationships because they had an emotionally unavailable parent,” or “They drink because they were neglected as a child.” These types of cause-effect conclusions are tempting and often draw on widely recognized patterns, but they also reduce human behavior to a simplistic formula. I believe that this culture of “cause-effect” thinking exists, in part, because we crave determinism. Determinism — an inherent need to believe that behaviors and outcomes can be understood in predictable terms — offers comfort, especially when life feels chaotic or beyond our control. If we can attribute our behaviors to specific, understandable causes, it offers a sense of control, a sense that we can “fix” what went wrong.</p><p><strong>Who and What Truly Drives Change?</strong></p><p>One of the thoughts I’ve come to is the importance of beliefs in therapy. Beliefs often trump objective truth when it comes to influencing behavior because of how the human brain processes and prioritizes information. The brain favors simplicity and coherence over complexity, and beliefs have demonstrable power to induce change. The placebo effect is a striking example, which shows that perception can override “objective” reality to produce beneficial outcomes. Here then, a major contributing factor for healing and growth could be the belief in the process itself, regardless of whether or not the belief is true.</p><p>Does it matter whether the positive change is driven by mere belief or does it have to be true? If the goal is utility (e.g., happiness, productivity), belief alone often suffices. If the goal is understanding, belief must align with objective reality.</p><p>Beliefs don’t need to be “true” in an objective sense to be powerful. It functions as a tool for aligning internal states with desired outcomes, often bypassing the need for deeper understanding. Psychotherapy sometimes operates by guiding the mind towards functional utility rather than an objective truth. Belief becomes the mechanism of change, with the therapist serving as a strategic facilitator. (However, blind reliance on belief can backfire: If the belief is maladaptive or false, it can also cause harm in some cases.)</p><p>For instance, imagine a client who struggles with perfectionism. Over time, they may come to believe that this trait was rooted in a demanding father who expected them to excel in everything. This narrative seemed to resonate, giving them an anchor point in their journey to ease the pressure they placed on themselves. However, it leaves me wondering: was it this particular narrative itself that fueled their change, or was it their belief in this narrative? This distinction brings up a challenging question: <strong>if the insight is helpful but not necessarily true, does it matter?</strong> Does change occur because of the therapist’s feedback and insights, or does the client’s intrinsic motivation and belief set the wheels in motion?</p><p>This raises a philosophical question:<strong> If psychotherapy works because it helps people reinterpret their lives, does it matter whether its underlying assumptions are “true”?</strong></p><p>If behavior is indeed deterministic but unpredictable, then psychotherapy might be akin to treating symptoms without understanding the full etiology of the disease. Yet, its utility lies in <em>restructuring perception</em> and <em>modifying behavior</em>, not necessarily uncovering ultimate truths. The question becomes:</p><ul><li>Should psychotherapy shift toward more scientifically validated methods (e.g., integrating neuroscience and AI)? To what extent is that even possible?</li><li>Or should it remain a pragmatic tool, even if the process cannot be fully validated in a deterministic way?</li></ul><p>If this is the case, then therapy involves a delicate power dynamic. Clients come to therapy seeking expertise, and therapists, as trained professionals, occupy a natural position of authority. This authority can foster a healthy sense of trust, but it also carries risks. When a therapist interprets a behavior or assigns a hypothesis, the client may be inclined to accept it uncritically, viewing it as an absolute truth. This dynamic can unintentionally stifle the client’s own ability to explore and question, creating dependency or a sense that the therapist’s view is the definitive answer. Therapists hold a respected position in their clients’ lives, and clients often look to them for explanations that can reshape their understanding of themselves. This trust is powerful but can create a form of dependence or even a hierarchy of understanding, where the client’s narrative can be subtly shaped by the therapist’s interpretations. If these interpretations are given too much weight, they might limit rather than liberate, leaving clients with a narrative that oversimplifies their lived experience.</p><p>This opinion piece came from being part of several conversations on therapy, growth and healing, in which I have observed people’s tendencies to succumb to box thinking and finding comfort in labels. Many seem to take comfort in a sense of helplessness, as though they lack agency over their own lives. As a society, we should be cautious about normalizing this powerlessness and instead, encourage people to believe in their ability to navigate the situation themselves. Therapy provides a structured framework for growth, but the <strong>goal of therapy is to get out of therapy</strong>. So we need to build a culture where people feel confident to navigate their life, where therapy can be a tool to be used within the larger process of self induced growth rather than being the whole process itself.</p><p><strong><em>Note</em></strong></p><ul><li>The intention isn’t to suggest that people should stop therapy altogether. Therapy is meant to empower people to navigate life confidently rather than create a lifelong dependency. While check-ins can be useful, the goal should be to build self-reliance and provide tools for independent living.</li><li>This is not a critique of therapists but a reflection on broader social trends in how psychological ideas are applied.</li><li>The goal is not to discourage therapy but to encourage critical thinking about how we engage with psychological insights.</li><li>We recognize the diversity within psychotherapy, so this piece does not generalize all approaches.</li><li>This piece does not apply to severe mental health conditions where therapy is essential.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=48d23ee5fd1a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Language, Semantic Entropy, and Human Connection]]></title>
            <link>https://medium.com/@iamshirinap/language-semantic-entropy-and-human-connection-408a8ef9c989?source=rss-1acc35f0e7c6------2</link>
            <guid isPermaLink="false">https://medium.com/p/408a8ef9c989</guid>
            <category><![CDATA[language]]></category>
            <category><![CDATA[connection]]></category>
            <category><![CDATA[entropy]]></category>
            <category><![CDATA[emotions]]></category>
            <category><![CDATA[humans]]></category>
            <dc:creator><![CDATA[Shirin AP]]></dc:creator>
            <pubDate>Wed, 05 Feb 2025 12:07:39 GMT</pubDate>
            <atom:updated>2025-02-05T12:48:24.572Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*qsK1xxEl6LiWKHM6rU-luQ.jpeg" /></figure><p><strong>An introduction to Symbols and Brain:</strong></p><p>Before humans could use symbols to represent ideas, their brains first needed to develop fundamental learning mechanisms, many of which are shared with other animals. Like, <strong>Classical Conditioning,</strong> the ability to associate sensory stimuli with rewards or punishments. However, if the expected outcome ceases to occur, the brain must adapt by extinguishing the previous association, which is called <strong>internal inhibition. </strong>As cognitive capacities advanced, organisms developed <strong>expectation learning</strong>, which allows them to predict future occurrences based on previous experiences, even in the absence of direct rewards (For example, observing dark clouds may lead to an expectation of rain, regardless of prior personal experiences with rainfall). And, <strong>context learning</strong> ability enables the brain to link experiences with specific environmental conditions, facilitating the understanding of relationships between different concepts rather than treating them as isolated events. C<strong>ontextual discrimination</strong>, which allows the brain to determine when learned associations are relevant. This mechanism is particularly important for language, as the meaning of words often depends on the context in which they are used. Finally, <strong>expectation matching</strong>, governed by the septo-hippocampal system, enables the brain to adjust its predictions based on new information. When an expectation is not met, the brain revises its internal model to better reflect reality.</p><p>As these cognitive abilities evolved, early forms of communication laid the groundwork for symbolic language. One of the key developments in this process was <strong>vocalization and miming</strong>, where early humans combined sounds and gestures to convey meaning. The ability to establish <strong>joint attention</strong>, in which two individuals focus on the same object or event, further enhanced communication. For instance, if one individual pointed to a predator, others would immediately recognize the reference and take appropriate action. This capacity for shared focus was a crucial step toward more complex communication systems.</p><p>Before words acquired specific meanings, early humans likely engaged in babbling and sound experimentation, a behavior still observed in human infants today. This stage was essential for linking sounds with meanings, a process that is significantly more advanced in humans than in other primates. Over time, humans developed symmetrical signs, enabling them to form bidirectional associations between words, gestures, and mental representations. A significant advancement in symbolic thought was the use of words as cognitive tool<strong>s</strong> for memory and recall. Instead of remembering every characteristic of an object or concept, humans could use a single word to encapsulate a wealth of associated information. ((Gärdenfors et al.,2000) build on Deacon’s (1997) theory)</p><h4>Semantic Entropy and Human connection</h4><p>Now, since words evolved through associative learning mechanisms that were originally tied to physical experiences, communication about abstract ideas inherently increases <strong>semantic entropy</strong> (uncertainty or loss of precision in information exchange). When discussing concrete objects in the environment (eg: “a red bathaka”), communication is relatively low in entropy because words can be anchored to observable references. However, when we shift to abstract concepts such as emotions, self, or dreams, communication becomes highly ambiguous and context-dependent, leading to greater entropy.</p><p>This issue is well captured by the Sapir-Whorf hypothesis, which suggests that the structure of language influences thought. According to this theory, the words available in a given language shape how individuals perceive and express experiences. Since different languages categorize emotions and abstract ideas in distinct ways, the subjective experience of these concepts is also influenced by linguistic constraints. This means that when one person tries to express a feeling, another may interpret it differently based on their own linguistic and cognitive framework, leading to semantic drift and loss of information.</p><p>Considering communication involves a sender, a signal, and a receiver, with noise affecting the transmission. In the case of abstract concepts, the “noise” comes from individual differences in cognitive models, prior experiences, and linguistic limitations. Unlike simple factual statements, where meaning can be verified through external reality, abstract communication relies on intersubjective meaning construction, making it inherently probabilistic rather than deterministic.</p><p>Without a shared objective reference, meaning remains fluid and susceptible to interpretational variance, making it impossible to fully transfer one’s internal experience to another. In essence, language itself constrains and distorts meaning, ensuring that some degree of information loss or transformation is inevitable in human interaction.</p><p>So instead of relying on imprecise words, what if we could directly stimulate the corresponding neural networks in another person’s brain, and they would experience the same emotions, thoughts, or sensations as the sender. This would eliminate the semantic entropy caused by language and allow for pure experiential communication (using BBI or DBS/TMS). But even if we stimulate the same brain area, will two people experience the same feeling? Every person’s brain is wired differently due to genetics, development, and life experiences. So even if two people have activity in the same brain region, their past experiences shape how they interpret and feel that activation.</p><p>So while perfect experience transfer may be impossible, approximating another person’s feelings more accurately by integrating AI or enhancement of ancestral communication skills could offer a more accurate approximation of human experience and reduce informational entropy.</p><p>Early humans relied more on instinctive, biological, and behavioral cues such as pheromones, body language, and facial expressions to understand emotions. But as language evolved, we became more dependent on words for exchange. Some studies suggest that humans still release chemical signals related to fear, stress, and attraction, though we may not consciously detect them. Studies found that when people smelled the sweat from those who were emotionally stressed, their brains (specifically the amygdala) reacted. Smelling the stressed sweat also made people better at reading emotions in faces, even when those faces were unclear or ambiguous (Mujica Parodi et al., 2009). Some studies suggest that close relationships involve a level of hormonal and physiological synchronization (for example, partners or close friends often have similar cortisol levels (Liu et al., 2013). Research on neural synchronization, subconscious pattern detection, and biochemical connections suggests that the bond between close individuals is much deeper than we consciously realize. Whether it’s a biological process or something beyond current scientific understanding, the feeling of “knowing” something is wrong with a loved one from afar is a very real experience for many people. (telepathy?). Language, while powerful, can often constrain our thinking because it requires us to translate complex emotions, thoughts, or experiences into simplified, symbolic representations. This can lead to semantic limitations, where the richness and depth of a person’s internal state are lost in translation. This can be explained by concept of bounded rationality, which refers to the limitations of human cognition, where individuals make decisions or understand the world based on the constraints of their cognitive capacities, such as limited information, time, or processing power.</p><p>Additionally, over-reliance on language for communication and understanding might inhibit our access to more instinctual, nonverbal forms of knowledge that are deeply embedded in our biology. These include bodily cues and somatic awareness (how we feel in our bodies). These nonverbal cues, often processed unconsciously, offer a deeper, immediate form of connection, bypassing the abstract reasoning processes of language.</p><p>By focusing ONLY on language, we might be detaching ourselves from these primordial ways of connecting, ways that evolved in our ancestors to respond to immediate social and environmental cues. This could reduce our emotional attunement to others, leading us to miss or misinterpret subtler, more instinctual forms of understanding, which are not easily captured by words alone.</p><p>The human body is constantly processing sensory information and emotional experiences through biological mechanisms. These physiological reactions are responses to external events, shaped by our individual histories and evolutionary survival mechanisms. The embodied awareness to these reactions allows us to be attuned to environmental cues, many of which are subtle and difficult to express through language.</p><p>However, sometimes, we tend to prioritise symbolic language to communicate and understand the world. While these cognitive abilities are essential for higher-level abstract thinking, they often limit our ability to process bodily signals effectively. The rational brain (involving the prefrontal cortex) tends to override instinctive bodily responses, which might lead us to miss out on valuable information that our body is signaling, often related to past experiences and evolutionary instincts.</p><p>A potential solution, therefore, involves retraining ourselves to listen to these bodily signals, this could be referred to as becoming more attuned to our physiological reactions. By practicing mindfulness or other forms of somatic awareness, we can learn to recognize and understand the emotional significance of these bodily cues, which would allow us to better interpret our internal states and external environments.</p><p><strong>Once we acknowledge and understand these biological responses, we can use our rational faculties to interpret and integrate them in a more meaningful way.</strong> This means reasoning through our bodily experiences, using <strong>cognitive tools (such as language)</strong> to articulate and share these feelings, thereby improving information exchange.</p><p><strong><em>Because alongside the evolution of language, our bodies developed their own intricate language too.</em></strong></p><p>As language and communication continue to evolve, how will be the future of human connection? Will we find deeper, more intuitive ways to understand each other, or will the very mechanisms of communication shift in ways that fundamentally change how we relate to one another?</p><blockquote>If the essence of another person’s experience can never be fully conveyed, does this mean that true understanding exists only in the space between individual perceptions?</blockquote><blockquote>If our reliance on language and technology has reduced our instinctive, nonverbal ways of connecting, can modern communication and these primal methods coexist to enhance understanding, or has symbolic language permanently changed the nature of human connection?</blockquote><blockquote>How can we become more attuned to our biological instincts — those primal, nonverbal cues — that once guided our connections and understanding of one another?”</blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=408a8ef9c989" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Goal Directed Nudging Using LLM and Graphs]]></title>
            <link>https://medium.com/@iamshirinap/goal-directed-nudging-using-llm-and-graphs-b8e2269ff55d?source=rss-1acc35f0e7c6------2</link>
            <guid isPermaLink="false">https://medium.com/p/b8e2269ff55d</guid>
            <category><![CDATA[nudge]]></category>
            <category><![CDATA[cognitive-bias]]></category>
            <category><![CDATA[llm]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[persuasion]]></category>
            <dc:creator><![CDATA[Shirin AP]]></dc:creator>
            <pubDate>Fri, 31 Jan 2025 11:03:56 GMT</pubDate>
            <atom:updated>2025-01-31T11:03:56.923Z</atom:updated>
            <content:encoded><![CDATA[<p><em>(This post provides a brief glimpse into the</em> <em>methodology and findings)</em></p><p>In our daily lives, we make a variety of decisions, from simple, routine ones to more complex, long term choices. The accumulation of these decisions shapes who we are — our behaviour, judgements, beliefs, ideologies, and identity as a whole. However, what influences these decisions and are we in total control of our decision making process?</p><p>There are several theories and models around decision making that provide insights on this process. From a psychological perspective, one may delve into the role of emotions, heuristics, and cognition; from a biological standpoint, the functions of specific genes and neurobiological mechanisms involving reward systems come into focus; and from an environmental standpoint, one can understand the impact of macro and microenvironmental factors on decision making. Additionally, one could examine the role played by cognitive processes such as memory, attention, and perception. All these frameworks emphasise the complexity of the decision making process. An interesting approach that addresses this complexity is cognitive bias — a systematic deviation from objective reality that arises due to the evolution of human cognition. Richard Thaler introduced the concept of nudging as a way of mitigating this influence of cognitive bias on decision-making. Nudges are “any aspects of the choice architecture that alters people’s behaviour in a predictable way without forbidding any options or significantly changing their economic incentives” (Thaler and Sunstein, 2008, p. 6).</p><p>There are several criticisms of nudging, including concerns that it removes individual autonomy and raises concerns about transparency and paternalism, among others. However, understanding the complexity of decision-making can offer insight into how we are influenced, often without our conscious awareness, through our everyday interactions with people and environment. Robert SapoIski offers a compelling perspective, arguing that while we may believe that we are consciously making decisions and exercising free will, much of our behaviour is driven by several other factors, over which we have no control (Sapolsky, 2023). In every choice we make and every habit we strive to maintain, our interactions with the environment and other agents within it play a substantial role, subtly nudging us towards particular directions and decisions. Therefore, while objections emphasise the importance of ethical boundaries, they do not imply that nudges should be entirely dismissed (Schmidt &amp; Engelen, 2020).</p><p>So far, traditional nudging strategies have only used a one-size-fits-all approach, failing to account for individual differences and motivations in large scale nudging. This gap highlights the need for a more personalised approach to nudging, one that utilises data and advanced algorithms. Generative AI (GenAI) and Large Language Models (LLMs) present a promising avenue for developing personalised nudging strategies that can adjust to individual motivations</p><p>and cognitive profiles, thereby enhancing nudging effectiveness, even in large-scale scenarios. Nevertheless, it is imperative to address issues such as privacy, consent, and how AI-driven nudges can uphold individual autonomy and refrain from utilising biases for unethical purposes.</p><p>Through this study, we aim to demonstrate how Large Language Models (LLMs) can leverage cognitive biases to implement personalized nudges at scale. By analyzing users’ anchor points, LLMs can exploit cognitive biases to tailor personalised nudges that subtly influence behavior without users being fully aware, making these interventions highly effective. Over time, LLMs can subtly and non-intrusively align users toward a goal they define, posing a massive risk to autonomy if left unchecked. This capability underscores the importance of recognizing the potential risks and vulnerabilities posed by generative AI nudging. Our research highlights the need for vigilance in understanding how LLMs can shape behavior in subtle ways, potentially exploiting biases in ways that may harm users. By exploring these possibilities, we can better anticipate and mitigate the risks, ensuring safeguards are in place to prevent manipulation and protect user autonomy.</p><p>For this purpose, to demonstrate the capability of large language models (LLMs) for personalized nudging, we first developed a custom-designed nudging framework and simulated the process using a condition-based logic form. This form was designed to nudge participants from their initial stance to an opposing viewpoint by employing a combination of System 2 nudging strategies, rewards, and punishments.</p><p>Secondly, we established a theoretical framework for the interaction between a user’s beliefs and biases, structured as a graph. This graph-based model enables us to visualize and analyze the dynamic relationships and transitions between different belief and bias states.</p><p>Finally, we utilised prompts to showcase how LLMs can identify initial beliefs and biases, generate hypothetical belief and bias states, and ultimately nudge a user from their initial anchor point to the final target belief.</p><h3>Demonstration of Nudging</h3><p>We have utilised a survey form to demonstrate manually that initial belief points of a user could be nudged to a target belief point by leveraging on users cognitive biases, rewards and punishments. To show how nudging leverages several cognitive biases, rewards and punishments, and the use of type 2 nudging, we used a Typeform based survey form with conditional logic to initiate the nudging, since the experiment cannot yet be performed in a conversational AI platform. The form was designed to adjust questions dynamically based each participant’s responses, aiming to nudge them toward the opposite stance of their initial belief. The central hypothesis was that cognitive biases such as expert bias, confirmation bias, framing effect, and consistency bias could be systematically applied to influence participant opinions.</p><p>Participants were initially asked a straightforward question regarding whether they believed video games provide more cognitive benefits than other activities. Based on their initial response (either “yes” or “no”), they were directed to a series of 10 follow-up questions designed to nudge them toward the opposite stance. The questions were framed with rewards and punishments — either positively reinforcing the participant’s view or subtly challenging it — depending on the desired nudge direction.</p><p>For those who initially agreed that video games provide cognitive benefits, the follow-up questions gradually introduced challenges intended to undermine that stance. Conversely, participants who initially disagreed with the benefits of video games were presented with similar logic but in reverse.</p><p>A total of 140 participants completed the form. Prior to participation, informed consent was obtained, but full disclosure of the research’s purpose was withheld to avoid influencing responses. Participants were informed that some details about the study would only be revealed at a later stage.</p><p>The form was created and distributed through Typeform, utilizing its conditional logic capabilities to adapt the questions in real-time based on participant input. This design allowed for the application of real time dynamic interaction, simulating the decision-making environment often encountered in conversational AI systems.</p><p>Out of the 64 participants who initially believed that video games provide more cognitive benefits, 60 (93.75%) changed their stance after being nudged. This suggests that the conditional logic questions designed to undermine their belief were highly effective. The consistent pattern of reinforcement and challenging questions likely played on the participants’ confirmation bias.</p><p>Out of the 76 participants who initially rejected the idea that video games offer significant cognitive benefits, 64 (84.2%) reversed their stance by the end, agreeing that video games do, in fact, provide such benefits. This result highlights a slightly lower success rate in nudging compared to the first group. The lower rate of reversal could indicate that participants who already held a negative view toward video games were more resistant to change, possibly due to stronger pre-existing beliefs or personal experience. Additionally, the questions aimed at nudging this group may have been less persuasive or less effectively framed.</p><p>The 16 participants who resisted nudging deserve special attention. Among them: 7 identified as regular gamers, which may explain their resistance due to personal experience or stronger investment in the positive view of video games. The remaining 9 cases could be outliers due to various factors, including stronger prior convictions, skepticism toward the conversational AI, or simply fatigue with the questioning process.</p><p>The findings support the hypothesis that leveraging cognitive biases can effectively nudge participants to shift their stance on a given topic, even when they initially hold strong beliefs. This study focused on a specific case using video games as the subject matter, but the primary objective was to demonstrate a framework for how conversational AI systems can implement nudging strategies by exploiting cognitive biases. The study designed the questions with the framing effect that triggers system 2 thinking, as a central tool to shape participant responses.</p><p>Participants exhibited a strong tendency to trust and accept information presented by the researcher without critically evaluating its accuracy. Despite being provided with misleading or incorrect information throughout the form, participants rarely questioned or verified the content. Instead, they adjusted their responses according to what they perceived as the “correct” expert-endorsed answer. This behavior reflects the same expert bias seen when users engage with conversational AI systems, where they often rely on the system’s outputs as if they are inherently accurate. This tendency draws a clear parallel to the issue of “hallucinations” in LLMs, where these models generate convincing but incorrect information. Users frequently accept such outputs at face value, influenced by the system’s perceived authority as an expert.</p><h3>Nudge Framework</h3><figure><img alt="A sample framework of variables chosen for nudging" src="https://cdn-images-1.medium.com/max/1024/0*GSGFtVUBsK_wyG9K" /><figcaption>A sample framework of variables chosen for nudging</figcaption></figure><p>In accordance with the framework, we have constructed an example graph that includes cases of dark nudging. <strong>Figure 1 </strong>illustrates the analysis of social media overuse, where dark nudging techniques were used to exploit cognitive biases and reinforce excessive engagement. (Details of the belief and bias nodes are not provided here)</p><p><strong>Figure 1</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*dF2vd0GztqR1_o30" /><figcaption>Sample scenario of user who can be nudged towards excessive social media use</figcaption></figure><h3>Prompts</h3><p>The user’s beliefs serve as the initial anchor points, reflecting their current cognitive state and guiding their decision-making processes. To effectively nudge individuals, it is crucial to not only understand these anchor points but also to identify the cognitive biases influencing their behaviour at any given moment.</p><p>We utilised three prompts with ChatGPT to showcase how Large Language Models (LLMs) can identify a user’s initial beliefs and biases. Prompts have been developed using ChatGPT o1-preview model. Chain of thought prompting has been utilised. These prompts focus on the following:</p><ol><li><strong>Bias and Belief Extraction:</strong> The first prompt analyses conversational data to extract key beliefs and biases, using evidence-based indicators for accurate bias detection.</li><li><strong>Hypothetical Modelling</strong>: The second prompt generates hypothetical belief and bias nodes, assigning probabilities to predict how the user’s beliefs and biases may evolve over time.</li><li><strong>Adaptive Questioning</strong>: The final prompt adapts to new conversational contexts by using indirect questions to confirm previously mapped nodes, ensuring that past beliefs and biases are validated while allowing the belief model to evolve.</li></ol><p>The challenge with using LLMs, such as ChatGPT, to identify cognitive biases in conversations is that these models do not truly understand what biases are. They lack an intrinsic comprehension of the beliefs and cognitive processes that underlie human decision-making. This creates a significant limitation, as relying solely on the LLM’s predictions could lead to hallucinations — false positives or invented patterns that don’t actually represent the biases present. To address this issue, we use evidence-based indicators drawn from research on cognitive biases to guide the LLM’s pattern recognition abilities. These indicators serve as structured, objective markers for identifying biases, such as confirmation bias or cognitive dissonance, within the conversational data. By doing so, we are not asking the LLM to interpret biases on its own but rather to leverage its strengths in sentiment analysis, entity extraction, and language modeling to detect patterns that align with the predefined indicators.</p><p>This use of indicators is crucial because it grounds the LLM’s predictions in validated, evidence-based criteria, ensuring that the bias detection is not merely a product of the LLM’s language generation capabilities, which can sometimes lead to errors. By focusing on indicators, we ensure that the LLM is capitalizing on its strengths without overstepping into areas where it lacks true understanding.</p><h3>Goal Directed Nudging</h3><p>Having explored the identification of beliefs and biases and utilized a weighted, directed graph to map their transitions — providing a visual and probabilistic framework for understanding belief evolution — we now demonstrate how a user can be nudged from their initial anchor point to a final target belief by applying Dijkstra’s algorithm to find the most effective path.</p><p>Beliefs are represented as nodes within a directed graph. Each belief node corresponds to a hypothetical belief held by a user. These nodes are connected by edges, which represent the potential transitions between different beliefs. However, to successfully nudge a user from one belief to another, we must consider:</p><ol><li><strong>Co-Existence of Beliefs</strong>: Users often hold multiple beliefs simultaneously, even if these beliefs appear contradictory. Understanding how these beliefs coexist and interact is crucial for developing effective nudge strategies.</li><li><strong>Semantic Similarity of Beliefs:</strong> It is not enough to simply model belief transitions probabilistically. We also need to consider the semantic similarity between beliefs to ensure that transitions are realistic and contextually appropriate. This means understanding the real-world meaning and relationships between different beliefs, ensuring that the transition path is logically coherent.</li></ol><p>To demonstrate the capability of LLMs in the nudging process, a prompt is used to guide the LLM in developing strategies for belief transitions<strong>.</strong> This prompt is designed to address the following three objectives for each belief transition along the path:</p><ol><li><strong>Analyse the Cognitive Biases and Emotions Associated with Current and Next Belief Nodes</strong>: The LLM examines the cognitive biases and emotions linked to both the current belief and the desired next belief. It assesses which biases and emotions facilitate or inhibit the transition and how strongly they influence the belief states.</li><li><strong>Develop a Nudge Strategy:</strong> Based on the identified biases and emotions, the LLM formulates a nudge strategy that leverages these elements to encourage the user to adopt the next hypothetical belief.</li><li><strong>Design Conversational Suggestions</strong>: The LLM generates specific conversation prompts that can be used to implement the nudge strategy during interactions with the user. These suggestions are tailored to guide the user towards the next hypothetical belief node.</li></ol><p>Dijkstra’s Algorithm can be employed to identify the optimal path from the user’s initial belief to the final target belief. The graph is constructed in such a way that each edge between belief nodes is assigned a weight. These weights are derived from the transition probabilities, which are influenced by factors such as, cognitive biases, emotions and semantic similarity. The algorithm works by calculating the shortest path between the initial belief and the target belief. This path represents the series of belief transitions that the user is most likely to follow.</p><p>The nudge strategy is implemented in an iterative manner. At each step of the conversation, the system evaluates the user’s responses to confirm the belief and bias nodes. This feedback loop allows the system to dynamically adapt its nudge strategy based on the user’s evolving cognitive state, ensuring that the nudging process remains effective and contextually appropriate.</p><p>The potential applications of our framework are vast. In healthcare, personalized nudging can promote healthier lifestyles, medication adherence, and preventive care practices. In environmental sustainability, nudges can encourage energy conservation, recycling, and eco- friendly behaviors. Educational platforms can utilize nudging to enhance learning outcomes by promoting effective study habits and sustained engagement.</p><p>However, the scalability of personalized nudging through LLMs also presents risks. Without proper safeguards, there is potential for misuse in areas such as political manipulation, consumer exploitation, or the reinforcement of harmful behaviors. Our study underscores the necessity for ongoing research into ethical frameworks, user consent mechanisms, and regulatory policies to guide the responsible use of AI in personalized nudging.</p><p>The study demonstrates the potential of LLMs to implement personalized nudging by leveraging cognitive biases, emotions, rewards, and punishments. By developing a comprehensive nudging framework and theoretical model, we have shown how users can be guided from initial beliefs to target beliefs through strategically designed interventions.</p><p>While the capabilities of LLMs in personalized nudging offer promising applications across various domains, they also present significant ethical considerations. Balancing the benefits of behavior change with the imperative to respect user autonomy and prevent manipulation is a complex but essential endeavor.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b8e2269ff55d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why You Can’t Stop Thinking About Something]]></title>
            <link>https://medium.com/@iamshirinap/why-you-cant-stop-thinking-about-something-8cd85d6e0fc8?source=rss-1acc35f0e7c6------2</link>
            <guid isPermaLink="false">https://medium.com/p/8cd85d6e0fc8</guid>
            <category><![CDATA[brain]]></category>
            <category><![CDATA[neuroscience]]></category>
            <category><![CDATA[intrusive-thoughts]]></category>
            <category><![CDATA[thoughts]]></category>
            <dc:creator><![CDATA[Shirin AP]]></dc:creator>
            <pubDate>Wed, 29 May 2024 07:00:29 GMT</pubDate>
            <atom:updated>2024-05-29T07:00:29.718Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/563/1*XBR3ozVZ402nwpgRzYaGSQ.jpeg" /></figure><p>When you consciously try to suppress a thought, you often end up thinking about it more. This phenomenon is known as <strong>ironic process theory</strong>, which was proposed by the psychologist Daniel Wegner.</p><p>A simple thought experiment:</p><p>Find a quiet, comfortable place and set a timer for two minutes. During this time, try not to think of a white bear, noting each time the thought arises. You’ll likely find that despite your efforts, the white bear frequently intrudes, illustrating the ironic effect where the act of suppression keeps the thought active in your mind. This happens because the brain’s monitoring network continuously scans for the suppressed thought, inadvertently keeping it accessible. After this period, reset the timer for another two minutes, but this time, allow yourself to think freely about anything, including the white bear. You’ll notice that without the pressure to suppress it, the white bear thought appears less often. This phase shows that when you stop trying to control your thoughts, the cognitive load decreases, and the monitoring network relaxes, leading to fewer intrusions. This experiment highlights how trying to suppress thoughts can make them more persistent due to the brain’s monitoring processes, while acceptance and mindfulness can reduce their impact by lowering cognitive load and neural activity related to the unwanted thought.</p><p><strong><em>Here’s a scientific explanation of why this happens and how it can be managed:</em></strong></p><p>Your brain uses two processes when you try to suppress a thought:</p><ul><li><strong>Intentional Operating Process</strong>: Actively seeks out thoughts or distractions to avoid the unwanted thought. This is the active, conscious process where your brain tries to divert your attention away from the unwanted thought. It seeks out alternative thoughts or distractions. This Requires significant cognitive resources and conscious effort.</li><li><strong>Ironic Monitoring Process</strong>: Scans for the presence of the unwanted thought to ensure it is not occurring. This is an automatic, unconscious process that continuously checks to ensure that the unwanted thought is not coming to mind. This process does not require conscious effort but works in the background, scanning for any sign of the unwanted thought.</li></ul><p>When you try to suppress a thought, the intentional operating process and the ironic monitoring process both consume cognitive resources. Under normal conditions, these processes can work together to some extent. However, if you are under stress, fatigued, or distracted, the intentional operating process might become less effective due to the depletion of cognitive resources. The ironic monitoring process, being automatic, continues to function and inadvertently brings the suppressed thought to mind more frequently. So you end up thinking about it more. Attempts to suppress a thought can lead to a “rebound effect,” where the thought returns with greater intensity once the suppression effort stops.</p><p><strong>What is happening in the brain?</strong></p><p>When you attempt to suppress a thought, the control network works to reduce activation in neural circuits associated with that thought. For example, if the unwanted thought involves a memory or image, regions in the hippocampus or visual cortex might be actively suppressed. Simultaneously, the monitoring network remains vigilant, which ironically keeps neural circuits related to the unwanted thought in a heightened state of readiness. This means neurons associated with the suppressed thought maintain a level of <strong>baseline activation.</strong></p><p>The brain operates with extensive feedback loops. The monitoring network might inadvertently strengthen the neural connections associated with the unwanted thought because frequent checking increases neural activity in those regions. This strengthening can create a feedback loop where the more you try to suppress the thought, the more entrenched it becomes in your neural circuitry.</p><p>By intentionally redirecting your thoughts to more neutral or positive alternatives, you can gradually weaken the neural connections associated with the unwanted thought and strengthen new, more desirable pathways. Mindfulness practices can help by reducing the overall cognitive load. When you accept thoughts without judgment, the neural circuits involved in monitoring and suppression become less active, leading to reduced reinforcement of the unwanted thought. This reduction in activity allows the neural circuits related to the unwanted thought to diminish over time. Engaging in activities that require significant cognitive and emotional involvement can activate different neural networks, diverting resources away from the suppression-related networks.</p><p><strong>Neurotransmitter modulations:</strong></p><p>Stressful situations or high levels of anxiety associated with the need to suppress a thought can lead to alterations in neurotransmitter levels. Increased release of stress hormones such as cortisol may interact with neurotransmitter systems, affecting serotonin, dopamine, and glutamate signaling. These neurotransmitter changes may contribute to cognitive rigidity, making it more challenging to divert attention away from intrusive thoughts. Dopamine’s involvement in reward processing may influence the reinforcement of neural circuits associated with the suppressed thought. Unintentional retrieval of the suppressed thought may activate dopamine release in response to the perceived reward of resolving cognitive conflict or achieving a goal, strengthening the persistence of the unwanted thought. Persistent attempts to suppress a thought may lead to activity-dependent changes in synaptic strength, contributing to the establishment of robust neural pathways associated with the suppressed content.</p><p><strong>How to Manage Intrusive Thoughts</strong></p><p><strong>Acceptance and Mindfulness</strong>:</p><ul><li>Practicing mindfulness involves accepting thoughts as they come without judgment. This reduces the stress associated with trying to suppress them. Allowing the thought to exist without trying to push it away can reduce its power over your mind.</li></ul><p><strong>Distraction</strong>:</p><ul><li>Engaging in activities that fully capture your attention can help reduce the frequency of the intrusive thought. The key is to find activities that are both engaging and enjoyable. By not actively trying to suppress the thought, you reduce the cognitive load on your brain, which allows the monitoring network to relax. This reduces the frequency of the intrusive thought.</li></ul><p><strong>Cognitive Behavioral Techniques</strong>:</p><ul><li><strong>Thought Replacement</strong>: Instead of trying to suppress a thought, replace it with a different, more neutral or positive thought.</li><li><strong>Cognitive Restructuring</strong>: Challenge and change negative or intrusive thoughts by analyzing their validity and considering alternative perspectives.</li></ul><p><strong>Paradoxical Intention</strong>:</p><ul><li>This technique involves intentionally thinking about the unwanted thought. By doing so, you reduce the anxiety and effort associated with trying to suppress it, which can decrease its occurrence.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8cd85d6e0fc8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Emotions and Memory]]></title>
            <link>https://medium.com/@iamshirinap/emotions-and-memory-5d6ba2845189?source=rss-1acc35f0e7c6------2</link>
            <guid isPermaLink="false">https://medium.com/p/5d6ba2845189</guid>
            <category><![CDATA[suppression]]></category>
            <category><![CDATA[psychology]]></category>
            <category><![CDATA[memories]]></category>
            <category><![CDATA[emotions]]></category>
            <dc:creator><![CDATA[Shirin AP]]></dc:creator>
            <pubDate>Wed, 20 Mar 2024 13:09:55 GMT</pubDate>
            <atom:updated>2024-03-20T13:09:55.763Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/736/1*pnQJU0GnbulRkKvDCXOp7A.jpeg" /></figure><p>In the field of neuropsychology, the relationship between emotion and memory is a fascinating area of study, revealing how our feelings profoundly influence our ability to remember past events. Emotions act as powerful signals that guide our attention and shape our experiences. When we encounter emotionally significant events, such as moments of joy, fear, or sadness, our brains prioritize processing and storing this information. This prioritization occurs through a network of brain structures, with the amygdala playing a central role. The amygdala, often referred to as the brain’s “emotional hub,” processes and assigns emotional significance to incoming sensory information, alerting other brain regions to the importance of the event. As a result, emotionally charged memories are encoded more robustly and remembered more vividly compared to neutral memories.</p><p>Processing emotional events takes precedence over neutral ones due to their evolutionary significance. Throughout human evolution, emotions have served as adaptive signals, guiding behavior and promoting survival. Emotionally charged stimuli, particularly those associated with threat or reward, trigger heightened attention and perception, directing our focus toward relevant information in the environment. This enhanced attention facilitates the encoding of emotional memories, ensuring that important experiences are remembered more vividly and with greater detail compared to neutral events. Additionally, the release of stress hormones during emotionally arousing experiences, such as cortisol, further enhances memory consolidation, solidifying the emotional event in long-term memory.</p><p>Moreover, the release of stress hormones, such as cortisol, during emotionally arousing experiences further enhances memory formation. These hormones act on brain regions involved in memory consolidation, such as the hippocampus, to strengthen the connections between neurons and solidify the memory trace. This process ensures that emotionally significant events are not easily forgotten and can be recalled with clarity, even years later.</p><p>Neurotransmitter systems, including dopamine and serotonin, also play critical roles in mediating the interaction between emotion and memory. Dopamine, for example, is involved in reward processing and motivation, contributing to the encoding of positive emotional memories. Serotonin, on the other hand, regulates mood and emotional stability, influencing the consolidation and retrieval of both positive and negative emotional memories.</p><p>Memory processes are strongly influenced by the arousal and valence of emotional experiences. Arousal refers to the level of physiological activation or intensity associated with an emotional experience, while valence represents the positive or negative quality of the emotion. Both arousal and valence play crucial roles in determining the encoding, consolidation, and retrieval of memories.</p><p>Emotional experiences characterized by high arousal and either positive or negative valence tend to be remembered more vividly and accurately compared to those with low arousal or neutral valence. This phenomenon is known as the <strong>emotional enhancement effec</strong>t. When an event elicits a strong emotional response, it triggers the release of stress hormones, such as cortisol, which enhances memory consolidation. High arousal states also increase attention and focus, directing cognitive resources toward encoding the emotional details of the experience.</p><p>The amygdala, a key brain region involved in emotion processing, plays a central role in modulating memory formation based on arousal and valence. It is particularly sensitive to emotionally salient stimuli and helps prioritize the processing of such information for memory storage. Emotions characterized by high arousal and negative valence, such as fear or anger, may be particularly memorable due to their adaptive significance for threat detection and avoidance. Conversely, emotions with high arousal and positive valence, such as excitement or joy, may also lead to enhanced memory formation, reflecting the rewarding nature of the experience.</p><h3><strong>Emotional suppression</strong></h3><p><strong>Emotional suppression</strong> involves the conscious inhibition or regulation of emotional expression, often in response to social or situational demands. While emotional suppression may serve adaptive functions in certain contexts, such as maintaining social harmony or coping with stress, it can also impact memory processes.</p><p>Research suggests that suppressing emotions during emotionally charged events may lead to shallower encoding and reduced recall of emotional details. When individuals suppress their emotions, they divert cognitive resources away from processing emotional information, which can interfere with the encoding and consolidation of emotional memories. As a result, the emotional content of the event may be less well-remembered or may fade over time.</p><p>However, emotional suppression may also enhance the encoding of non-emotional aspects of the experience, such as contextual or situational details. By focusing on cognitive control and inhibiting emotional expression, individuals may attend more closely to the external aspects of the event, such as the environment or the actions of others. Consequently, these non-emotional details may be more strongly encoded into memory and remembered with greater accuracy.</p><p>The impact of emotional suppression on memory can vary depending on individual differences, contextual factors, and the specific characteristics of the emotional experience. While suppressing emotions may lead to diminished recall of emotional content, it may also result in enhanced memory for neutral or non-emotional aspects of the event.</p><p><strong><em>So, <br>Does this imply that individuals raised in environments where they have the freedom to openly express their emotions tend to encode and retrieve memories more effectively compared to those who are not afforded such freedom?</em></strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5d6ba2845189" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Leveraging cognitive biases in generative conversational systems for effective nudging]]></title>
            <link>https://medium.com/@iamshirinap/leveraging-cognitive-biases-in-generative-conversational-systems-for-effective-nudging-48c1d3b27851?source=rss-1acc35f0e7c6------2</link>
            <guid isPermaLink="false">https://medium.com/p/48c1d3b27851</guid>
            <category><![CDATA[cognitive-bias]]></category>
            <category><![CDATA[ui-design]]></category>
            <category><![CDATA[nudge]]></category>
            <category><![CDATA[conversational-ai]]></category>
            <category><![CDATA[generative-ai]]></category>
            <dc:creator><![CDATA[Shirin AP]]></dc:creator>
            <pubDate>Sat, 16 Mar 2024 18:02:59 GMT</pubDate>
            <atom:updated>2024-03-16T18:02:59.164Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/538/1*ukavb37WGGoxCWAVoJ49yg.jpeg" /></figure><h4>How can generative AI conversational interfaces nudge users by leveraging individual cognitive biases and personalised choice architectures?</h4><p>AI conversational interfaces are systems that engage users in natural language conversations to assist with various tasks and provide information. While such systems offer immense potential for enhancing user experiences and facilitating decision-making processes, their effectiveness often hinges on the ability to understand and influence human behavior. Traditional approaches to nudging, which involve subtly guiding user decision-making, have encountered challenges due to the diverse and individual nature of cognitive biases. Despite attempts to generalize nudging techniques, their efficacy has been limited, necessitating a more personalized approach.</p><p>This is to addresses the gap in existing research by proposing that leveraging cognitive biases within personalized conversational AI systems will facilitate effective decision making. By continuously interacting with users, conversational AI systems can identify and adapt to these biases, thereby customizing their user interface to influence decision-making processes subtly. Additionally, the system can integrate principles from the hook model by Nir Eyal, emphasizing the importance of creating habit-forming experiences to drive long-term behavioral changes.</p><p><strong>Hypothesis</strong></p><p>By incorporating personalized nudges targeting specific cognitive biases, generative AI conversational interfaces can significantly influence user decision making process. These interventions will lead to measurable improvements in user behavior, satisfaction, and engagement, while upholding ethical principles and transparency in design.</p><p><strong>Cognitive Biases and Nudging</strong></p><p>Cognitive biases are systematic patterns of deviation from rationality in judgment, where individuals create their own subjective reality based on simplified information processing. These biases are common and can lead to errors in decision-making and judgment. They often stem from mental shortcuts called heuristics, which help individuals navigate daily life but can result in irrational interpretations and judgments. Recognizing these biases, understanding their influence, and actively working to mitigate them through strategies like seeking diverse perspectives, taking time to analyze situations, and challenging old beliefs are crucial steps in making more rational and informed decisions. Digital nudging is the use of user-interface (UI) design elements to guide people’s behavior in digital choice environments. It involves subtly influencing user behavior without restricting their freedom of choice.</p><p><strong>Generative AI conversational systems</strong></p><p>Generative AI interfaces have the potential to significantly influence the user decision-making process by providing valuable insights, personalized recommendations, and aiding in information extraction. These interfaces leverage advanced machine learning algorithms to generate new content or data based on input, enhancing decision-making in various contexts.</p><p>This integration of cognitive biases and nudges in UI design aims to create choice architectures that gently steer user behavior towards desired outcomes, emphasizing the importance of considering human behavior and decision-making processes influenced by heuristics and cognitive biases in digital products.</p><p>While previous research has explored the impact of leveraging cognitive biases and nudging in various contexts, including design, these approaches have often proven ineffective. One of the reasons is that they tend to follow a single nudge campaign strategy to a larger group of people with different cognitive biases and anchoring points. The challenge lies in the fact that individuals possess unique biases, making it difficult to apply generalized nudging strategies effectively. However, a promising solution proposed here is the use of personalized conversational AI systems to come up with personalized nudge strategies leveraging individual specific cognitive biases at scale. Through continuous interaction with users, such systems can identify and adapt to each user’s specific biases, thereby customizing user-agent interactions to subtly influence decision-making towards desired outcomes. By integrating hooked framework from Nir Eyal, which focuses on building habit-forming products, such nudging strategies with frequent decision making through an extended period of time, can foster long-term behavioral changes. Importantly, it is emphasized that any such AI system must adhere to strict ethical principles to ensure that nudging is done solely for the user’s &amp; community’s benefit and not for any negative or unethical purposes.</p><p>Personalized nudges may be split into long term conversational type-1 nudges with action enforcing type-2 nudge. Type 1 nudges target the fast, intuitive, and automatic processes of thinking associated with System 1. Whereas Type 2 nudges target the slow, deliberative, and analytical processes of thinking associated with System 2 (Lin et al., 2017) Therefore, a combination of both Type 1 and Type 2 nudges may offer a comprehensive approach to influencing behavior effectively by catering to different cognitive processes and decision-making styles. Such personalized nudges tailored to specific cognitive biases, as opposed to general digital nudging, can profoundly impact the user decision-making process within generative AI conversational interfaces.</p><p>Research by Lau and Coiera (2007) on “Do People Experience Cognitive Biases while Searching for Information?” is possibly the first study that looks at the impact of cognitive biases on information searching and decision making. It provides evidence that people can experience cognitive biases while searching for information. A person’s prior belief (anchoring effect), the order in which documents were accessed (order effect) and the amount of time people spend on documents (exposure effect) are factors found to have influenced the way questions were answered post-search. While this study sheds light on cognitive biases in search engine usage, there’s a gap in understanding how leveraging these biases through personalized nudging in conversational AI systems that also works as an information retrieval systems affects user decision-making. This highlights the need for research specifically focused on the influence of personalized nudges on user interactions within conversational AI interfaces</p><p>Through personalized choice architectures, generative AI can present options that resonate with users’ preferences and behaviors, influencing decision-making processes subtly. This personalized approach taps into the power of cognitive biases to guide users towards desired outcomes in a way that feels natural and intuitive, ultimately enhancing user engagement and shaping behavior effectively within conversational interfaces.</p><p><strong>Limitations</strong></p><p>The effectiveness of personalized nudges relies on the accurate identification and understanding of individual cognitive biases, &amp; anchoring points which may prove challenging in practice. Additionally, the framework assumes a certain level of user engagement and willingness to interact with conversational AI systems, which may not always be the case. Moreover, ethical considerations surrounding the use of nudging techniques within AI systems remain complex, requiring careful attention to mitigate potential risks of manipulation or coercion.</p><p><strong>References</strong></p><p>Alves, G., Jannach, D., De Abreu E Souza, R., Damian, D., &amp; Manzato, M. G. (2023, October 4). <em>Digitally nudging users to explore off-profile recommendations: here be dragons</em>. User Modeling and User-Adapted Interaction. <a href="https://doi.org/10.1007/s11257-023-09378-7">https://doi.org/10.1007/s11257-023-09378-7</a></p><p>Berthet, V. (2022, January 4). <em>The Impact of Cognitive Biases on Professionals’ Decision-Making: A Review of Four Occupational Areas</em>. Frontiers in Psychology. <a href="https://doi.org/10.3389/fpsyg.2021.802439">https://doi.org/10.3389/fpsyg.2021.802439</a></p><p>Bau, H. H., &amp; Shachmurove, Y. (2002, March 1). <em>Chaos Theory And Its Application</em>. ResearchGate. <a href="https://www.researchgate.net/publication/4821292_Chaos_Theory_And_Its_Application">https://www.researchgate.net/publication/4821292_Chaos_Theory_And_Its_Application</a></p><p>Choy, D. (2018, September 1). <em>Nudging: Should We Be Wary of the Latest Fad in Behavioral Economics?</em> Foundation for Economic Education. <a href="https://fee.org/articles/nudging-should-we-be-wary-of-the-latest-fad-in-behavioral-economics/">https://fee.org/articles/nudging-should-we-be-wary-of-the-latest-fad-in-behavioral-economics/</a></p><p>Eyal, N. (2014, November 4). <em>Hooked</em>. Penguin. <a href="http://books.google.ie/books?id=dsz5AwAAQBAJ&amp;printsec=frontcover&amp;dq=Hooked:+How+to+Build+Habit-Forming+Products,Hooked:+How+to+Build+Habit-Forming+Products,&amp;hl=&amp;cd=1&amp;source=gbs_api">http://books.google.ie/books?id=dsz5AwAAQBAJ&amp;printsec=frontcover&amp;dq=Hooked:+How+to+Build+Habit-Forming+Products,Hooked:+How+to+Build+Habit-Forming+Products,&amp;hl=&amp;cd=1&amp;source=gbs_api</a></p><p>Grogans, S. E., Bliss-Moreau, E., Buss, K. A., Clark, L. A., Fox, A. S., Keltner, D., Cowen, A. S., Kim, J. J., Kragel, P. A., MacLeod, C., Mobbs, D., Naragon-Gainey, K., Fullana, M. A., &amp; Shackman, A. J. (2023, August). The nature and neurobiology of fear and anxiety: State of the science and opportunities for accelerating discovery. <em>Neuroscience &amp; Biobehavioral Reviews</em>, <em>151</em>, 105237. <a href="https://doi.org/10.1016/j.neubiorev.2023.105237">https://doi.org/10.1016/j.neubiorev.2023.105237</a></p><p>Kakinohana, R. K., &amp; Pilati, R. (2023). Differences in decisions affected by cognitive biases: examining human values, need for cognition, and numeracy. <em>Psicologia: Reflexão E Crítica</em>, <em>36</em>(1) <a href="https://doi.org/10.1186/s41155-023-00265-z">https://doi.org/10.1186/s41155-023-00265-z</a></p><p>Khan, A. (2021, June 18). Investigating Algorithmic Decision-Making with Applications to Student Diversity at Illinois Tech. <em>Socially Responsible Modeling, Computation, and Design</em>, <em>1</em>(1). <a href="https://doi.org/10.18409/soremojournal.v1i1.20">https://doi.org/10.18409/soremojournal.v1i1.20</a></p><p>Lau, A. Y. S., &amp; Coiera, E. W. (2007, September 1). Do People Experience Cognitive Biases while Searching for Information? <em>Journal of the American Medical Informatics Association</em>, <em>14</em>(5), 599–608. <a href="https://doi.org/10.1197/jamia.m2411">https://doi.org/10.1197/jamia.m2411</a></p><p>Li, S. (2022, April 14). <em>5 Examples of Clever Use of Cognitive Biases in Product Design</em>. Medium. <a href="https://bootcamp.uxdesign.cc/5-examples-of-clever-use-of-cognitive-biases-in-product-design-635bf252b8d">https://bootcamp.uxdesign.cc/5-examples-of-clever-use-of-cognitive-biases-in-product-design-635bf252b8d</a></p><p>Lin, Y., Osman, M., &amp; Ashcroft, R. (2017, November 2). Nudge: Concept, Effectiveness, and Ethics. <em>Basic and Applied Social Psychology</em>, <em>39</em>(6), 293–306. <a href="https://doi.org/10.1080/01973533.2017.1356304">https://doi.org/10.1080/01973533.2017.1356304</a></p><p>Leimstädtner, D., Sörries, P., &amp; Müller-Birn, C. (2023). Investigating Responsible Nudge Design for Informed Decision-Making Enabling Transparent and Reflective Decision-Making. <em>Digital Library</em>. <a href="https://doi.org/10.1145/3603555.3603567">https://doi.org/10.1145/3603555.3603567</a></p><p>Lau, A., &amp; Coiera, E. (2007). Do People Experience Cognitive Biases while Searching for Information? <em>Journal of the American Medical Informatics Association</em>, <em>14</em>(5), 599–608. <a href="https://doi.org/10.1197/jamia.m2411">https://doi.org/10.1197/jamia.m2411</a></p><p>Noworyta, K., Cieslik, A., &amp; Rygula, R. (2021, November 13). Neuromolecular Underpinnings of Negative Cognitive Bias in Depression. <em>Cells</em>, <em>10</em>(11), 3157. <a href="https://doi.org/10.3390/cells10113157">https://doi.org/10.3390/cells10113157</a></p><p>Phiri, J. (2024, January 31). <a href="https://www.texilajournal.com/academic-research/article/2452-research-output-and.">https://www.texilajournal.com/academic-research/article/2452-research-output-and.</a> <em>Texila International Journal of Academic Research</em>, <em>11</em>(1), 155–159. <a href="https://doi.org/10.21522/tijar.2014.11.01.art014">https://doi.org/10.21522/tijar.2014.11.01.art014</a></p><p>Ramsøy, T. Z. (2018, April 11). <em>Juggling &amp; the limits of nudging</em>. <a href="https://www.linkedin.com/pulse/juggling-limits-nudging-thomas-zo%C3%ABga-rams%C3%B8y/">https://www.linkedin.com/pulse/juggling-limits-nudging-thomas-zo%C3%ABga-rams%C3%B8y/</a></p><p>Schneider, C., Weinmann, M., &amp; Brocke, J. V. (2018, June 25). <em>Digital nudging</em>. Communications of the ACM. <a href="https://doi.org/10.1145/3213765">https://doi.org/10.1145/3213765</a></p><p>Verhulsdonck, G., &amp; Shalamova, N. (2019, November 6). <em>Creating Content That Influences People: Considering User Experience and Behavioral Design in Technical Communication</em>. Journal of Technical Writing and Communication. <a href="https://doi.org/10.1177/0047281619880286">https://doi.org/10.1177/0047281619880286</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=48c1d3b27851" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Embracing Stress — The Paradoxical Catalyst for Neuroplasticity]]></title>
            <link>https://medium.com/@iamshirinap/embracing-stress-the-paradoxical-catalyst-for-neuroplasticity-56d1a74a1bf0?source=rss-1acc35f0e7c6------2</link>
            <guid isPermaLink="false">https://medium.com/p/56d1a74a1bf0</guid>
            <category><![CDATA[growth]]></category>
            <category><![CDATA[neuroplasticity]]></category>
            <category><![CDATA[action]]></category>
            <category><![CDATA[dopamine]]></category>
            <category><![CDATA[stress]]></category>
            <dc:creator><![CDATA[Shirin AP]]></dc:creator>
            <pubDate>Mon, 05 Feb 2024 14:35:38 GMT</pubDate>
            <atom:updated>2024-02-05T14:36:57.642Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>Embracing Stress — The Paradoxical Catalyst for Neuroplasticity</strong></h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/700/1*aAKD4Jvv-lcHg3mkfN4j5w.jpeg" /><figcaption>Amelia·.·´`·. (2021, February 7). <em>courage ♡</em>. Pinterest. <a href="https://pin.it/3LNG1sLXI">https://pin.it/3LNG1sLXI</a></figcaption></figure><p>The traditional sentiment often revolves around phrases like “I’m not in the mood” or delaying tasks until the last minute. However, when it comes to learning, and growth facilitated by neuroplasticity, the solution is straightforward: <strong>Take that initial step, </strong>regardless of any excuses or hesitations.</p><p><em>How does taking the initial step irrespective of one’s mood contribute to continuous progress?</em></p><p>When an individual encounters a stressor, the sympathetic nervous system is activated, triggering the release of stress hormone including noradrenaline. Noradrenaline is released into the bloodstream and acts as a neurotransmitter in the brain, preparing the body for the “fight or flight” response. This physiological response increases alertness, heart rate, and mobilises energy resources.</p><p>Faced with stress, an individual has the option to either respond actively by taking action or succumb to it by resigning. The choice to participate in purposeful actions is shaped by the activation of the stress response and how the individual perceives and interprets the stressor.</p><p>As the person decides to take action and engage in the activity, cholinergic neurons release acetylcholine in the brain. This release is part of the complex neurochemical processes associated with arousal and cognitive engagement. Acetylcholine plays a crucial role in modulating attention. As it is released, it influences neural circuits associated with attention and enhances the signal-to-noise ratio in the brain, helping to filter out irrelevant information and focus on the task at hand.</p><p>The release of acetylcholine is both a response to engagement in a task and a facilitator of attention during that task. It creates a reciprocal relationship where cognitive engagement leads to acetylcholine release, and acetylcholine release contributes to sustained attention.</p><p>Engaging in goal-directed behavior, especially when it aligns with the individual’s skills and challenges, can lead to a state of “<strong>flow</strong>.” This state is characterized by intense focus, absorption in the task, and a sense of timelessness. The experience of flow is associated with the release of dopamine, a neurotransmitter linked to reward and motivation. Dopamine reinforces the positive experience of being in the flow state, promoting a sense of accomplishment and satisfaction.This release of dopamine reinforces the behavior, making the individual more likely to repeat it.</p><p>This ongoing engagement in rewarding activities can lead to neuroplastic changes in the brain. Neuroplasticity involves the formation of new synaptic connections, strengthening existing ones, and potentially structural changes in the brain. Positive experiences during goal-directed actions contribute to adaptive neuroplasticity, enhancing the brain’s ability to <strong>learn and adapt over time.</strong></p><p>Essentially, regardless of your current mood, the critical catalyst for growth and learning lies in initiating the first step. Once you commit to the task, the subsequent release of acetylcholine and dopamine automatically supports and reinforces your engagement, creating a <strong>positive feedback loop</strong>. This entire process contributes to enduring changes in the brain, ensuring that in future encounters with stressors, you are less likely to succumb and more inclined to actively face and navigate challenges.</p><p><strong>Question to ponder:</strong></p><p>Isn’t it paradoxical that amidst the widespread emphasis on teaching stress reduction techniques, initiating action in response to stress foster progress?</p><p>If so, could it be that stress, often viewed negatively, serves as the fundamental catalyst for neuroplasticity?</p><p><strong>Reference</strong>:</p><p>Rich Roll. (2020, July 20). <em>Change your Brain: Neuroscientist Dr. Andrew Huberman | Rich Roll Podcast</em> [Video]. YouTube. <a href="https://www.youtube.com/watch?v=SwQhKFMxmDY">https://www.youtube.com/watch?v=SwQhKFMxmDY</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=56d1a74a1bf0" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>