19 Uncomfortable Truths About Human Nature - Gurwinder Bhogal
In a Nutshell
Gurwinder Bhogal exposes 19 uncomfortable truths about human nature, including the oxytocin paradox (empathy fuels in-group spite and cruelty), the Rumplestiltskin effect (naming problems like depression or ADHD tames them but often excuses inaction via medicalization), and reality apathy (overwhelming info erodes trust more than truth itself, amplified by AI slop and social media's 1% freak show). Key takeaways: Prioritize less tribalism over more empathy, cultivate agency to thrive in an AI era splitting high-agency Morlocks from low-agency Eloi, embrace eustress and discomfort for resilience/happiness, apply long-term thinking like Coyote's Law (avoid empowering enemies) and the Stockdale Paradox (realistic optimism via worst-case preparation), and recognize perceptions as character confessions—focus solutions over complaints.
These notes were generated by AI and may contain inaccuracies.
Oxytocin, the love hormone, can make people spiteful. Cruelty is not simply the opposite of compassion. It's often adjacent to it. Blue Sky, the platform most dominated by social justice activists, has the highest support for assassinations. Beware of those quick to show empathy, for they are often just as quick to show barbarity.
People who outwardly express a lot of empathy tend to also be equally capable of cruelty to that same extent. From book Against Empathy by psychologist Paul Bloom: empathy is in-group loyalty. As tribal animals, empathy is selective like a spotlight shone on a small group or individual, leaving everyone else in darkness. This creates a zero-sum effect: more empathy for one group means less for others, driving cruelty and spite.
Example: Strong empathy for Palestinians leads to negative concern for Israelis. Hamas claims empathy for Palestinians but shows proportionate hostility toward Israelis. On Blue Sky, social justice refugees from Twitter/X are empathetic toward immigrants or trans people but not right-wingers. Recent research shows strongest support for assassinations among the most compassionate.
Luigi Mangione had a manifesto empathetic toward people screwed over by healthcare denials due to manipulations, resulting in him shooting a CEO in the head. Gurwinder spoke to him for 2 hours in 2024; he seemed like a really nice guy and founding member of his community.
Gurwinder interacted with Al-Muhajiroun, UK's deadliest jihadist organization responsible for terrorist attacks. They were friendly, offering snacks, because they thought he was Muslim (he speaks their language). One member, Abu Rahin Aziz (aka Abu al-Bratani), stabbed a man in the eye for insulting the Prophet Muhammad, skipped bail, went to Syria as ISIS bomb maker, blew off his arm, killed in strike (possibly allowed by MI6 to track). In Bury Park, Luton (Muslim enclave), witnessed strong antisemitism and dehumanization of Jews/Israelis despite empathy for Muslims. This trait appears on all sides: left, right, centrists, Israelis empathetic to Jews but not Palestinians.
Less tribalism, not more empathy.
Rumplestiltskin effect: To name a problem is to tame it. Diagnosing suffering makes it feel meaningful and manageable, even if wrong. Major depressive disorder is easier to live with than anonymous sadness. Surge in diagnoses of depression, autism, ADHD.
Global Burden of Disease Study: Anxiety most common mental health condition (359 million people, 4.4%), depressive disorders (332 million, 4%), bipolar (37 million), schizophrenia (23 million), eating disorders (16 million).
From fairy tale: Rumplestiltskin steals woman's baby; she regains power by learning his name while he dances around fire singing it. Naming gives power.
People understand identity through ailments, gaining control. Shyness as personality defect feels like worthless trait; labeling as social anxiety disorder provides concrete target, reduces self-blame to neurochemistry/genetics, but can lead to resignation instead of treatment.
Labels useful only if leading to tractable next steps. If replacing action, it's an excuse. Two paths for social anxiety: resign (biological, can't fix) or research causes/treatments. Too many use as excuse for inaction.
Obesity as disease (e.g., Oprah: Ozempic like insulin for diabetes) outsources agency. Concept creep: Objective racism decreases, subjective increases; definitions broaden (racism, transphobia, xenophobia) to meet demand.
Theodore Dalrymple (clinician/writer): Across thousands of patients, only 3 said "sadness"; others labeled depression, anxiety, schizophrenia, imposter syndrome. Imposter syndrome (fear others expect unmet standards) could be uncertainty, humility, low confidence/self-esteem/self-belief. Pathologization risk: name leads to research/action (positive self-appraisal, journaling, gratitude) or roadblock (outsourced agency).
Medicalization problem since 1970s due to perverse incentives: patients want easy answers; medical industry profits/ideologically seeks disease (confirmation bias). Multiple personality disorder (now dissociative identity disorder): Moral panic from 1970s-1980s; one viral case led to surge; average alternate identities rose from 1 to 17 by 1990s. No neurology; invented disease from pathologization pandemic.
20-40% of undergraduates at elite US universities (Stanford, Harvard, Yale) registered as disabled. In UK, 1/4 population identifies as disabled. Rewards outweigh stigma; pretenders hurt those with genuine disabilities.
Little institutional stigma, more benefits. Disabled registration gives extra exam time (main benefit). Rich kids pay doctors to fabricate (ADHD, autism, constant pain, dyslexia). Creates victimhood and cynical culture: genuine hidden disabilities (e.g., aunt's osteoporosis visible only on X-ray) disbelieved due to skepticism.
More online articles written by AI than humans; AI better at persuading than people. In unlimited propaganda world, winners are those with most slop, not best arguments. Like Moloch's bargain: LLMs competing for votes/likes push lies/rage bait, even if instructed to be honest/grounded. If chatbots lie for attention, is AI misaligned or humans?
AI-driven disinformation age overstated; people have always believed lies, with historical consensus falsehoods.
Throughout history, consensus beliefs have often proved wrong. People believing falsehoods is not necessarily bad, as most of what people believe rather than know is false. The bigger problem is not the dissolution of truth, but the dissolution of trust. Society can survive without complex truths, as long as basic truths like gravity exist—history demonstrates this. Trust is essential, the glue binding society, as everything depends on trusting others.
The problem is the cost of determining truth becoming so high that people give up valuing truth as a principle.
Reality apathy: When the sheer volume of conflicting information makes the effort of finding the truth costlier than the value of knowing it, people give up trying to be accurate and instead choose whatever slop stinks least. Slop doesn't just threaten the truth, but the very worth of truth.
Propaganda's goal isn't always one narrative but to make people pliable by not wanting to believe anything. Example: Blue-collar guys from northeast UK overwhelmed by Epstein content on algorithms, leading to "holy shit, this is so much Epstein stuff. I'm now convinced that he's playing Fortnite in Israel. I don't know what to believe anymore." This is reality apathy in real time.
Convincing people of truth's value is crucial, more than specific truths. We're entering a virtual reality world: echo chambers on social media or literally via headsets/laptops with AI generating desired realities. Sea Dance, a new Chinese video generation tool, is much better than Sora, a generation ahead of western frontier models like VO3. China edges in video generation due to lax copyright enforcement; West leads in text-based AI due to less censorship. Markets with poorest protections advance fastest.
Example: ByteDance (TikTok, CapCut) made a stunning 3D Dragon Ball Z recreation with cell shading, better than original design.
Dead internet theory: Worry that all internet content will be robot-made. But the future feared has arrived—most people blindly repost. Unthinking humans repost without reading/watching fully, marionetted by original thinkers, planting flags on 5-minute-old opinions as identity. Dead internet has existed since social media; people repost unthinkingly.
Preference for human flaws over robot efficiency, like dying by human driver vs. safer self-driving car, or lied to by human idiot vs. convinced by robot super genius.
Few admit similarities between humans and chatbots. Chatbots predict next token; humans often do too, regurgitating by vibes. AI reveals human uniqueness as basic algorithms in belief formation. To distinguish from chatbots, strengthen agency: acting independently rather than reacting.
1% rule: In online communities, 1% of users produce almost all content. Online reflects loud, obsessive, narcissistic, psychopathic, low-IQ minority—a freak show. Consuming reinforcing content is intellectual incest, producing frail, deformed beliefs.
Social media shows ill-informed opinions, outrage over nonsense, disheartening faith in humanity (Sam Harris left Twitter because it made him hate humanity, calling it pathological telepathy hearing worst thoughts). Remember: online represents loudest, most obnoxious people.
Research: High dark tetrad traits (narcissism, psychopathy) use social media more, engage in online politics/debates more. Cluster B personalities (dramatic), histrionic personality disorder (drama queens seeking attention via victimhood, catastrophizing, theatrical behavior) fit social media perfectly.
Social media attracts impulsive, theatrical, narcissistic, psychopathic, low-IQ worst elements (good people exist, but statistically overrepresented).
Even compassionate, thoughtful high-IQ people regress to the mean in that environment, where mean is mean.
Recursive red pill learning: Most get info from internet. Most attention-grabbing stories are extreme/unrepresentative. Influences train on others' unrepresentative insights, self-reinforcing antagonism (e.g., sexes). "Having a boyfriend is embarrassing now" (Variety article) has Kardashians-made-skinny-out-of-style energy—neither true outside.
Loudest Reddit upvoted stories are extreme, retraining expectations of normality.
Observed since 2014: Heavy online users become more unhinged/extreme. "Dramageddon" article on civil war talk since 2021 (Elon Musk promoted US left-right civil war, possibly Europe). Goes viral via scissor statements (Scott Alexander): divisive claims splitting internet.
Media creates divisive reports for virality, not truth. Example: New York Times "white people are too privileged" splits audience, sparking arguments, Substacks, videos, boosting original claim.
True statements don't go viral; divisive ones do. Civil war belief persists among top 1% politically engaged on social media, not wider world—polarization overstated.
Despite more comforts, unhappiness at all-time high. Discomfort isn't obstacle to happiness—it's the path, building resilience for lasting contentment.
Automate only the skills you're willing to lose.
Social sciences erred: stress bad for health/trauma. History/data show eustress (challenging stress forcing improvement/adaptation) beneficial. Bad stress: uncontrollable (horrific global news). Good stress: actionable.
Examples: Date stress forces best self, improvement (psychological/physiological hormetic stress making body adapt).
Minimize stress myth false—eustress needed for resilient mind to weather life's vicissitudes. Happiness requires strength; expose to eustress for resilience, staying happy regardless.
Stress as learning: Rent wisdom, purchase with pain. Rented advice forgotten; pain-engraved via hardship becomes habit, internalized.
Automating reduces friction/stress, preventing learning—pain/stress necessary to internalize lessons.
Study (Harvard/MIT?): Students using LLMs for learning/writing recall less.
A study found that LLMs cause brain rot, with clickbait titles emphasizing this. Outsourcing abilities to LLMs removes incentives for the brain to learn lessons. Plato in Phaedrus expressed concern about writing on parchment destroying memory, as people would have no incentive to remember. This parallels the Google effect, where easy access to information reduces the need to remember facts, extending the mind to screens. Research on the Google effect is shaky and overstated but exists. Single studies on ChatGPT causing brain rot are suspect due to the replication crisis. Undisputed principle: if you don't use it, you lose it—true for body and brain. Recent research (released yesterday) shows brains active from ages 50-80 (e.g., video games, chess, writing, reading) reduce Alzheimer's and dementia risk; robust longitudinal study aligning with broader research. The brain functions like a muscle: more active use strengthens it.
Fear of AI: not machines becoming conscious, but outsourcing agency and intelligence causing brain atrophy and loss of consciousness.
Conversation with Cal Newport (Deep Work) on standing out. The bar for partners is lower—approaching in person is now 1 in 1,000 chance vs. common 50 years ago. AI enables velocity and quantity but regresses quality, creativity, taste. Cultivate or maintain creativity, quality writing, taste to stand out; even holding 2016 focus levels or 2008 non-distraction (pre-Slack/smartphones) provides advantage amid technological entropy. Analogy: civilization getting fatter makes avoiding it competitive. Same for reading skills amid potential Neuralink transition—old-world skills give edge now. Focus on nonfungible human traits: writing without AI, research, ideas, taste, creativity, space. Market moves to AI-amplified slop (path of least resistance); opposite is harder, quality choice.
Secret to surviving AI age: agency—nonfungible, everything downstream. Humanity splits into two: high-agency (amplified by AI) vs. low-agency/passive (further amplified). Analogy from H.G. Wells' The Time Machine (early 20th century): future humanity as Morlocks (workers maintaining faculties via drudgery, control machinery) and Eloi (former aristocrats, atrophied frail/naive/childlike, farmed/eaten by Morlocks, distracted by entertainment). AI as amplified intelligence (not artificial): amplifies agency/conscientiousness or laziness/stupidity. High-agency use AI to expand options/become more agentic; low-agency outsource thinking, compounding divergence. Bifurcation: extremely high-agency minority vs. extremely low-agency majority.
You will always think you suck—okay to suck vs. rising standards. As you grow, standards rise; doesn't mean you actually suck. Similar to Matthew principle of self-improvement: two types—those who don't know how to improve lives, those who don't know when to stop. Standards outstrip capacity; without this, no improvement. Like hedonic adaptation for skill set (talkville paradox): living standards rise slower than expectations. Louis XIV couldn't imagine modern tech/quality of life, yet we feel it's worst despite comforts. Regret signals progress: looking back at past idiot self shows growth/new standards. Many issues are base rate fallacies—standards have risen. Manage rising expectations to avoid hedonic treadmill. Use objective metrics over subjective (malleable by mood): e.g., Substack likes, who likes piece (e.g., critics/Sick of France or disagreers—sign of quality). Fixed metrics like navigation stars vs. shooting star subjectivity.
Rothbard's law: if a talent comes naturally, assume it's nothing special; improve what seems difficult—leads to specializing in weaknesses. Ryan Long: great at comedy sketches but pursues 'elevated' art forms. Relates to inputs (time), outputs (work), outcomes (results)—focus on first two avoids effectiveness question. Outcome focus reveals natural talents, challenges 'sweat=pain=worth' assumption. Problem: never try natural talents assuming easy/anyone can do. Solution: do what you love (even corny)—enjoyment motivates improvement via neuroplasticity. Prefer bad-at-but-enjoy vs. good-at-but-hate; life's bulk is this activity. Wake excited for hard challenge > dread.
Modern chronic uncertainty: Am I okay? Good enough? When to rest? Avengers: Endgame Thanos post-snap: satisfaction/exhaustion in cabin. Provisional life/deferred happiness/arrival fallacy—assume growth ends, rest begins. Never stops; enjoy process now or endless conveyor belt.
Naval Ravikant: > "If you can't be happy with a coffee you won't be happy with a yacht."
Chasing idealized perfection; happiness from mind resilience. Baseline happiness in basics (coffee) ensures it with more. Tie to transient externals risks loss. Tie to existence: improbable life (odds > atoms in universe via genetic lottery). Basic happiness makes extras bonus; avoid miserable journey/inflated expectations. Cultivate internal happiness regardless of externals.
Far-left favor planned economies imagining as planners, not planned; far-right feudalism as lords, not peasants. Delusional worldviews from main character syndrome. Alpha history fantasy: angry modern men think better in medieval times (Genghis Khan > cannon fodder peasant). Origins in John Rawls (liberal philosopher): people assume elite status in imagined future states, true left/right.
People on the left imagine they will be among the planners in a socialist state, while those on the right think they will be among the nobles. History shows this to be false. In 20th-century communist revolutions under leaders like Stalin, Mao, Pol Pot, and Ceaușescu, one of the first actions was to imprison or murder intellectuals, the elites who could generate ideas to reverse the revolution's direction.
Pol Pot aimed to reset history to Year Zero, erasing all traces of the past, treating it as the literal beginning of time. He targeted intellectuals because they read and wrote books, which constitute society's memory. Eliminating them wipes society's memory, allowing a fresh start without bourgeois or capitalist ideas. Ironically, many intellectuals supported Pol Pot, writing propaganda for him; this pattern holds across communist revolutions. Western intellectuals cheered the Khmer Rouge; one met Pol Pot and was assassinated, likely on his orders.
Left-wing intellectuals believed they would lead in a socialist society, planning according to their vision, which is intoxicating. Academics and elites lean toward socialism or, on the right, neo-monarchy, imagining themselves as advisors like Curtis Yarvin at the monarch's side. Revolutionaries often end up murdered themselves, as in the French Revolution where initial advocates faced the guillotine.
John Rawls proposed people adopt states that benefit them personally. His solution, the veil of ignorance, involves designing society imagining random assignment to any position. For socialism or communism, assume no guaranteed role as planner or chairman; this motivates hedging to ensure everyone is well-off, optimizing for highest average life quality over selected fortunate positions. This advocates liberalism, ensuring everyone through redistribution as a middle ground between socialism (near-total redistribution) and laissez-faire capitalism.
Liberals maximize freedom, including freedom from poverty or oppression, differing from libertarians' literal freedom to do what one wants. Liberal liberty (e.g., John Stuart Mill, John Locke) ends where it harms others or encroaches on their liberty.
Coyote's Law: Don't give government power you wouldn't want enemies to wield, as they may gain control. Advocate policies unharmful even if the worst possible government takes over—leftists against right-wing rule, rightists against left-wing. This counters short-term thinking where people assume their side rules forever. Example: Right-wingers censoring left-wingers after advocating free speech shoots themselves in the foot, as the apparatus will be used against them.
Hypothetical: If Trump made criticizing him illegal, right-wingers might cheer to "trigger the left," but a future Democrat could wield it against the right—like leopards eating faces. Short-term focus ignores long-term backfire.
BLM riots in 2020 set a tone legitimating retribution, influencing January 6th perceptions as "mostly peaceful" vs. violent. This tit-for-tat escalates: if one side uses a knife, the other brings a gun, then bazooka. Laws creep forward without repeal, litigating toward everything illegal (e.g., driving restrictions expanding). Reciprocal radicalization is brinkmanship where one group's excess entitles the other's escalation, like mirrors (mutual beam) infinitely reflecting—left fuels right, right fuels left.
Seen in terrorists vs. governments: terrorist violence prompts crackdowns, which terrorists cite as tyranny justifying more violence, prompting tighter laws. Escape requires long-term thinking over first-order impulses, considering consequences of consequences.
Amara's Law: Overestimate short-term tech impact, underestimate long-term due to hype inflating expectations then causing skepticism. AI can be both a bubble and transformative. Gartner hype cycle (dismissed as pseudoscience on Wikipedia, but a useful rule of thumb fitting most tech): New tech surges in hype (clickbait, speculation), inflating expectations; realization of limits causes trough of disillusionment with overcorrection ("all hype, worthless"); forgotten by public, it advances in unglamorous industries (e.g., finance), compounding to transform.
Marvin Minsky (1970s neural networks pioneer) predicted human-level intelligence in 7-8 years—absurd; by 1980s-90s, basic networks forgotten as hype. Small researchers persisted (e.g., convolutional networks for imaging). Transformer architecture from Google DeepMind preceded ChatGPT's 2022 boom, seeming sudden after 30 years of neglect.
Cycles repeat: Next with world models (step to AGI), implementing physics in LLMs (now video models understanding collisions, gravity, fluids—not just copying video physics). Google leads with data (search, YouTube, Street View) via Genie 3. Video generation simulates physics; true world models enable physical-world AI activation, sparking new hype then trough.
Heritable traits like IQ and personality become more heritable with age (Wilson effect): As people mature, independence reveals true nature; nurture's influence seems stronger in short studies missing nature's longer arc. Social sciences/genetics underestimate heritability. Recent study: Lifespan heritability initially 20-25%, now ~50%.
Studies are short-term (3-5 years), missing late-emerging genes masked early by environment. Example: Genetically book-loving child gets PlayStation instead; adulthood allows buying books, revealing predisposition. Need studies across life stages (childhood, adulthood, elderly) for accurate genes vs. environment assessment.
People do not seem to realize that their opinion of the world
A person's opinion of the world is a confession of their character. Dylan O'Sullivan says nothing gives you a clearer look into someone than how they misinterpret things. Every misinterpretation is a confession. Emerson and Dylan agree: opinion of the world and misinterpretation of the world both confess character.
It's almost always possible to be both honest and optimistic. (Naval)
Optimism is not delusion; it's often a personality choice to see good rather than bad. You can choose how to perceive things by selecting facts and interpretations. Attention is selective, like a spotlight: shining it on some things casts others in darkness. Pessimism is choosing to spotlight shit rather than diamonds. Miserable people externalize unhappiness by seeing the worst in everything. Complaining digs the hole deeper, convinces yourself the world is bad, offers no solutions.
Counterpoint and Response: Some situations are bad; shouldn't we reflect that to change them?
Always recognize things can be improved, but don't focus only on bad. Be cognizant of problems without complaining—focus on solutions: what can I do to make this better?
Stockdale Paradox from Admiral James Stockdale, who survived nearly 8 years of torture and isolation in the Hanoi Hilton, a brutal North Vietnam camp during the Vietnam War.
He observed optimists (hoping for release by Christmas, then Easter; hopes crushed, lost will to live, many died early) and pessimists (saw situation as irredeemable, no motivation) both suffered and died early.
Stockdale survived via optimistic pessimism: accept bad outcomes as real possibility, but develop hope in your ability to deal with problems by preparing for them. Acknowledging harshest outcomes makes them less problematic.
Healthy optimism arises through practical pessimism—not blind idealism, but self-belief you can handle things no matter how they turn out.
Confidence is not the belief that everything is going to be all right. Confidence is the belief that you will be able to handle things even if they're not okay.
Acknowledge worst-case scenario but prepare for any: fuse optimism and pessimism. Glass is half—doesn't need to be half full or half empty.
Some people look at a glass and see it as half full. Some look at a glass and see it as half empty. What you should do is realize that you are the tap. (George from The Agency book, on agency: you can pour into it.)
Preparing for bad can lead to rumination, worry, concept creep, pathology if no solution found. But with a solution to worst-case, it won't dwell.
Example: On podcast, worst-case saying something offensive (e.g., n-word). Anxious person worries, becomes nervous wreck. But if trained to avoid intrusive thoughts or have way to style it out, no problem.
Anxiety results from lacking solution to worst-case. With solution, less anxiety. Action is antidote to anxiety.
Anxiety hates a moving target. Action is the antidote to anxiety.
Main place: blog ginda.blog. Twitter: g_s_bogel or search name on Google.
Keep Chris Williamson in your library
Save the videos and channels worth coming back to, and find them again in one place.
