Day 7: How emotions control what you think, who you trust, and what you remember — Understanding Emotion

← All Days

The final day. Emotions are not passive. They actively distort memory, attention, and judgement. Anxious people cannot focus because anxiety is a mood that keeps the survival system running in the background. Frightened people cannot do anything else because fear is an emotion that locks every resource onto one thing. Empathy is not a mirror response. It is feeling what others feel in your body. Psychopaths have the lowest form. Highly sensitive people have the highest. The people we need are those who can manage it.

Emotions actively shape three core cognitive functions: memory, attention, and judgement. They are not passive feelings. They change how you think
Anxiety is a mood. Fear is an emotion. This distinction explains why anxious people cannot focus (persistent background noise) while frightened people cannot do anything else (acute lockdown)
Emotional events are remembered far more vividly and durably than neutral ones. The amygdala tags emotional experiences for priority storage. This is why trauma memories are so hard to shake
We judge other people based on our own emotional state at the time. Meet someone when you are happy, you rate them more positively. Meet the same person when you are anxious, you rate them as less trustworthy. The person did not change. You did
Empathy is not a mirror response. It is feeling what others feel in your own body. Psychopaths have the lowest form, highly sensitive people have the highest. The people we need are those who can manage empathy: feel it genuinely but still think clearly enough to act

Mood vs Emotion: Why This Distinction Matters for Attention

A mood is a slow burn. An emotion is an explosion. They hijack your focus in completely different ways

Day 5 introduced the difference between mood and emotion. Happiness is a mood. Fear is an emotion. But the distinction goes further than that. It explains why different emotional states destroy your focus in different ways
Anxiety is a mood. It is low-level, persistent, and does not have a single clear trigger. When you are anxious, your body is in a mild survival state for hours or even days. The sympathetic nervous system stays slightly activated. Cortisol stays elevated. Your brain is constantly scanning for threats that may not even exist. This is why anxious people find it hard to concentrate on work, reading, or conversations. The survival scanner is running in the background, eating up cognitive resources like a program draining your battery
Fear is an emotion. It is acute, intense, and tied to a specific trigger. When you are frightened, every cognitive resource locks onto the threat. Your visual field narrows. Your hearing sharpens. Your body prepares to run or fight. There is no bandwidth left for anything else. A frightened person does not struggle to focus. They hyper-focus, but only on the source of fear. Everything else disappears
Here is a concrete example. Imagine you have a presentation tomorrow. If you are anxious about it, you will spend the entire evening unable to enjoy dinner, unable to read a book, unable to sleep properly. The anxiety sits in the background and poisons everything. But if a fire alarm goes off during the presentation itself, you feel fear. You stop talking. You stop thinking about the slides. Every part of you focuses on getting out of the building. The presentation does not exist anymore
This maps perfectly to the survival purposes from Day 3. Fear is built for immediate threats: a predator, a falling rock, a fire. It needs total focus to survive. Anxiety is built for ongoing uncertain threats: a rival tribe nearby, food running low, social tension in the group. It needs persistent vigilance, not total focus. Evolution built two different attention-hijacking systems because the threats they address are structurally different
This also explains the dopamine machines from Day 5. Social media generates chronic low-level anxiety (am I missing out? did someone respond? what are people saying about me?). It is not fear. It is mood-level anxiety that keeps the survival scanner running all day. The phone never triggers a fear response. It triggers something worse: a mood that quietly eats your ability to focus on anything else

Emotions and Memory: Why You Remember the Terrible Days

The amygdala acts as an emotional highlighter. Events tagged with strong emotion get priority storage in long-term memory

You can probably remember exactly where you were when you heard terrible news. The room, the lighting, who told you, what you were wearing. But you cannot remember what you had for lunch last Tuesday. This is not random. Emotional events are stored differently in the brain
The amygdala is the key player here. When you experience something emotionally intense, the amygdala activates and essentially tells the hippocampus: 'this one matters, store it properly.' The result is what psychologists call a flashbulb memory: vivid, detailed, and durable. People who lived through major disasters, attacks, or personal tragedies often report near-photographic recall of the moment
This is an evolutionary advantage. If a particular location, animal, or situation caused you intense fear, you need to remember it in detail so you can avoid it next time. A vague memory of 'something bad happened near water' is less useful than a precise memory of exactly which river bend had the crocodile. The emotional tagging system exists to prioritise survival-relevant information
But there is a cost. Traumatic memories are also stored with this priority system. This is why people with PTSD have intrusive, vivid, involuntary memories of the traumatic event. The amygdala tagged the experience as maximally important, and now the memory replays with full emotional intensity. The system that was built to help you remember the crocodile is now replaying a car accident on loop
This connects to the exposure therapy discussion from Day 5. The reason exposure therapy works for PTSD is that it gradually weakens the emotional tag attached to the memory. You revisit the context without the bad thing happening, and over time the amygdala stops flagging it as dangerous. The memory remains, but the emotional charge fades. You remember the event, but it stops hijacking your body
Positive emotions also enhance memory, but less dramatically. You remember your wedding day or the birth of a child more vividly than a random Tuesday. Joy tags memories for storage too, but fear and distress are stronger tags. Evolution prioritised remembering threats over remembering rewards. Forgetting a good fruit tree is inconvenient. Forgetting a predator's den is fatal
There is a secondary effect too. Your current mood biases what memories you can access. When you are sad, sad memories surface more easily. When you are happy, happy memories come to mind faster. This is called mood-congruent memory. It means depression is partly self-reinforcing: the depressed mood makes it easier to recall depressing events, which deepens the depressed mood. A feedback loop

Emotions and Judgement: We Do Not See People as They Are

We see people through the filter of whatever we are feeling at the time. And first impressions are almost impossible to override

This one is unsettling. We think we judge people based on who they are. We do not. We judge them based on how we feel when we encounter them
If you meet someone at a party where you are relaxed and having a good time, you are more likely to rate them as warm, funny, and trustworthy. If you meet the exact same person on a day when you are stressed and irritable, you rate them as cold, boring, or suspicious. The person is identical. Your emotional state is the variable. This is called affect-as-information: your brain uses your current feeling as data about the external world
Here is why this is dangerous. First impressions are extraordinarily sticky. Once you form an initial judgement of someone, all future information gets filtered through that judgement. This is the halo effect. If your first impression is positive (because you were in a good mood), you will interpret their future behaviour charitably. If your first impression is negative (because you were anxious or angry), you will interpret the same behaviour as confirmation that they are not trustworthy
A real example. Imagine a job interview. The interviewer had a terrible morning: traffic, spilled coffee, argument with a partner. The candidate walks in and gives a perfectly good answer. But the interviewer's emotional state colours their perception. The answer seems 'okay but not great.' Another day, same interviewer, good mood: the same answer seems 'confident and well-structured.' This is not hypothetical. Studies on judicial sentencing have shown that judges give harsher sentences before lunch (when hungry and irritable) and more lenient ones after eating. The crime is the same. The judge's emotional state is the variable
This also explains why conflicts escalate. If you are already angry and someone does something mildly annoying, your anger makes you interpret the action as deliberately hostile. You respond aggressively. They get angry in return. Now both people are judging each other through anger, and the situation spirals. Neither person is seeing the other clearly. They are seeing reflections of their own emotional state
The practical takeaway: never make important decisions about people when you are in a strong emotional state. Do not fire someone when you are angry. Do not hire someone because they made you laugh. Do not end a relationship during a depressive episode. The emotion is real, but the judgement it produces is not reliable

Empathy Is Not a Mirror. It Is Feeling What Others Feel

Most people think empathy means understanding someone's situation. It does not. It means actually feeling their pain, their joy, their fear. And not everyone has the same capacity for it

There is a common misunderstanding about empathy. People think it is a cognitive exercise: you see someone in pain, you understand they are in pain, you respond appropriately. That is not empathy. That is sympathy, or at best cognitive perspective-taking. Real empathy is feeling the other person's emotion in your own body. Their grief sits in your chest. Their anxiety tightens your stomach. You do not observe their suffering. You share it
This is not a metaphor. There is real neural machinery behind it. When you watch someone get hurt, parts of your own pain matrix activate. Not the parts that register the physical sensation, but the parts that register the emotional distress. Your brain literally runs a partial simulation of what the other person is going through. You feel a shadow of their pain
Psychopaths have the lowest form of empathy. This is not a moral judgement. It is a neurological observation. Psychopaths can read facial expressions. They can tell you what emotion someone is displaying. They have the cognitive ability to identify emotions in others. But they do not feel them. The simulation does not run. They see your pain the way you see a weather forecast: informational, not felt. This is why psychopaths can hurt people without hesitation. The brake that stops most of us from causing suffering, which is feeling a version of that suffering ourselves, does not exist for them
Highly sensitive people have the highest form of empathy. At the other end of the spectrum are people who feel too much. They walk into a room and absorb the mood. Someone else's sadness becomes their sadness. Someone else's anxiety becomes their anxiety. This is not a choice. It is the empathy system running at maximum gain. Highly sensitive people are often emotionally exhausted because they are constantly processing not just their own feelings but everyone else's too
The people we actually need are not at either extreme. We need people who can feel what others feel but also manage that feeling. The psychopath is useless for cooperation because they cannot be moved by your distress. The overwhelmed empath is useless for leadership because they drown in everyone's feelings and cannot act. The sweet spot is someone who feels your pain genuinely but can still think clearly enough to help. That is managed empathy. It is a skill, not just a trait
This connects to the guilt discussion from Day 4. Guilt is an empathy-adjacent emotion. You feel guilty because you can simulate the harm you caused to someone else. If you could not feel their pain at all, like a psychopath, you would feel no guilt. And if you felt their pain too intensely, like a highly sensitive person, the guilt would be paralysing. The functional middle ground is feeling enough guilt to correct your behaviour without being destroyed by it
I want to go much deeper into this. I want to re-read 'The Selfish Gene' by Richard Dawkins with this lens. Dawkins argues that genes are the fundamental unit of selection, not organisms. If that is true, then empathy is a gene-level strategy: your genes benefit when you cooperate with others who share them, and empathy is the mechanism that makes cooperation feel good rather than being a cold calculation. The whole point of these emotions, guilt, love, empathy, might be gene-level engineering to make us cooperate. I want to understand whether empathy is a selfish strategy dressed up as selflessness. And more broadly, I want to use Dawkins to tie together everything from this experiment: why emotions exist, why empathy exists, and what all of it says about the point of being alive

What Is Next: Emotions and Machines

Two experiments I want to run next. Both connect this emotion research to the thing I actually build: neural networks

1. Do LLMs understand emotions? After seven days of studying how humans feel, I want to flip the question. Do large language models have anything resembling emotional understanding? Not whether they feel emotions, but whether they have learned the structure of emotions from training data. There is a paper from Anthropic I want to read: 'Emotion Concepts and Function'. If LLMs can model emotional states well enough to predict human behaviour, that tells us something interesting about whether emotions are computable or whether they require the biological hardware we spent seven days studying
2. Can you tune emotions into a neural network? This is the experiment I am most excited about. If emotions in the brain come down to neurotransmitter levels and activation patterns in specific regions, can we do something analogous in a trained neural network? Take an already-trained LLM and tweak specific neurons or groups of neurons to make it more sad, more empathetic, more cheerful, more anxious. Not through prompting. Through direct weight modification. Like turning up the serotonin dial but for a transformer. If this works, it would be direct evidence that emotional behaviour is an emergent property of network structure, not something that requires biological substrate. And if it does not work, that tells us something too: maybe emotions need the body, the hormones, the hydraulics that Freud talked about. Either way, it should be a fun experiment to run