Tarun Kumar Mahato

Every coherent aesthetic in the world is a particular combination of underlying primitives (typography, color, spacing, motion, imagery, voice, interaction). There are not infinite styles to template; there is a finite set of primitives that combine into infinite styles. The product's job is to identify which primitives the user responds to and compose a coherent output from them. The compositional rules prevent incoherent combinations.

The phrase "people lack taste" is almost true and slightly wrong. People don't lack taste; they lack resolution. They can tell good from bad when they see it side by side. They can't generate good from a blank page.

This is the crucial insight. Hantavirus kills brutally. Its death rate can reach 50%, far higher than COVID. But killing your host quickly is evolutionarily counterproductive for a pandemic pathogen. COVID spread silently for days before symptoms appeared. That is what makes something pandemic-shaped.

Most loneliness is not done to us. It is partly maintained by us. We hide the real thing out of fear. We show people the polished version. We test them with small vulnerabilities, and when they fail the test, we say "see, nobody really understands me" and retreat further. It becomes self-sealing. The loneliness protects you from the risk of being truly seen and rejected. Which is the deepest rejection there is.

True. But the more unsettling thing is that humanity doesn't need capital incentives to cause harm. We do plenty of that for free. Tribalism. Pride. Revenge. Ideology. The tweet implies we would be better without market incentives distorting our nature. But our nature was already complicated before capital existed. Money just gives our existing tendencies better infrastructure. The fickleness was always there. We just keep inventing more efficient systems to express it.

unbelievable how benefiting or harming humanity always agrees with immediate capital incentives. incredible thing, that humanity, very fickle

2.5 years. Still going. Still being called a "situation."

A number stopped meaning anything somewhere around 10,000. That's the thing nobody will say out loud. When Gaza crossed 10,000 dead, then 30,000, then 50,000, then 75,000 -- something in the human brain filed it under "ongoing situation" and moved on. We didn't stop caring. We just ran out of the cognitive machinery to care at that scale. Evolution never prepared us for grief at this resolution. And the people who run wars know this. They have always known this. Here's what's actually happening right now in 2026: A "ceasefire" exists on paper. Israel controls 60% of Gaza and keeps expanding. Aid organizations have been banned. The UN says four of the five legal definitions of genocide have been met. 270 journalists are dead -- more than in two decades of Vietnam. A family tree of 132 relatives was wiped out in a single airstrike on an apartment building. Meanwhile the war metastasized. US-Israeli strikes on Iran. Khamenei dead. The Strait of Hormuz effectively closed. Europe's jet fuel running short. The region is on fire in ways that will take decades to cool. And we are all having arguments on the internet about whose suffering counts more. This is what war does that nobody talks about enough: It forces everyone to pick a shape for their grief. You are allowed to mourn Israeli families who lost children on October 7. You are allowed to mourn Palestinian families who have lost everything since. The world keeps telling you to choose one. The world is wrong. The people who actually suffer wars are almost never the people who start them. Hamas leadership survived. Netanyahu survived. The generals survived. The weapons manufacturers made money. The people who paid -- really paid -- were a 6-year-old girl in Gaza City who never learned to read. A 24-year-old Israeli at a music festival who just wanted to dance. They have nothing to do with each other's politics. They were just born into the wrong geography at the wrong moment. The deepest thing I can say is this: We have built a world where it is easier to justify 75,000 deaths than to find a political arrangement that lets two groups of traumatized people live near each other. That is not a statement about Israel. Or Palestine. Or America. Or Iran. That is a statement about what we have collectively decided is acceptable. Every government that sent weapons and called it "support." Every government that issued statements and called it "concern." Every algorithm that ranked the outrage content and called it "engagement." Every person who picked a tribe and stopped being able to see the other side's dead as fully human. All of it contributed to a world where this was possible. I don't know how it ends. Nobody does. The machinery of this conflict is older than most nations currently debating it. But I know this: The children did not choose this. Not in Gaza. Not in Israel. Not in the Lebanese villages. Not anywhere. And the day we stop being able to say that clearly -- regardless of flag, regardless of which side we were raised to see as human -- is the day we've lost something we won't easily get back.

Brilliant article. Genuinely. The data is real, the historical precedents hold, and the lump-of-labor point is correct. But let's name what this actually is. a16z has billions invested in AI being good for humanity. And just published 4000 words arguing AI is good for humanity. That doesn't make you wrong. But it makes you a stakeholder protecting a position. Not a neutral observer. The people most harmed by this transition aren't reading your Substack. They don't have retraining options or LinkedIn networks. They have specific skills in specific places that are becoming cheaper to replace than they were two years ago. Your article gives them 26 charts and "history says it works out." History says it works out for the economy. It's largely silent on the specific human being surviving the specific gap. The footnote almost admitted why this article exists. Public fear about AI threatens regulation. A well written optimistic piece from a trusted source is cheaper than lobbying and more effective. One question your article never asks. When the transition comes, and it's already coming, who bears the cost? And what are you personally doing about that with the billions you're making from the technology causing the disruption? That answer isn't in the article. And its absence is the most honest thing about it.

The majority of new jobs created since 1940 didn’t even exist in 1940. There is no fixed "lump of labor". Again and again, new technologies create new jobs. a16z's David George dismantles the "AI job apocalypse" myth: a16z.news/p/the-ai-job-a…

The Jabalpur photo that made millions cry was AI generated. But here is what nobody is saying. Your grief was real. The manipulation was real. And the nine actual people who died are still waiting for that same grief. We live in a world now where suffering can be aesthetically manufactured to trigger empathy on demand. While real suffering, unglamorous and uncomposed, drowns quietly in the noise. You didn't cry because you are weak. You cried because someone weaponized the most human thing about you. Be angry about that. Then go find the real names. Gladvin's sister. His mother. His four year old nephew. Still missing. No AI composed their absence for your feed

The most underrated thing about getting rich is that the first dollar is harder than the first million. Not motivationally. Psychologically. Because the first dollar requires you to believe a stranger will pay you for something you made. That belief is the hardest thing in entrepreneurship. Not the code. Not the marketing. Not the strategy. The belief that you have something worth paying for.

@elevenreader I would love to join the challenge... audiobooks

Announcing the Audiobooks Creator Challenge. Create. Publish. Win. Win AirPods Max & credits to publish your full book.

The "War on Terror" created more terrorists than it killed.

America spent $8 trillion on the War on Terror. With that money, you could have: Eliminated global poverty Provided clean water to every human on earth Funded education for every child for decades Instead... bombs. Drones. Prisons. Torture. And at the end: more enemies than when you started.

"Terrorist" is a word the powerful use for violence they don't like. When America bombs a wedding in Yemen... it's "collateral damage." When someone bombs a market in New York... it's "terrorism." The violence is the same. The bodies are the same. The grief is the same. The only difference is who's doing it. That doesn't mean terrorism isn't real. It is. It's violence against civilians to create fear for political goals. But notice: states do this too. They just don't call it terrorism. They call it war. Policy. Necessity.

If you have whatever is important to you, you can consider yourself to be wealthy.

If you can convince a majority that they're under threat from a minority... they vote as a bloc. They consolidate. They stop asking questions about jobs, inflation, water, electricity. They vote on identity. On fear.

Schopenhauer said life is suffering. We want, we lack, we strive, we fail. Even when we get what we want, we just start wanting something else. Desire is infinite. Satisfaction is temporary. The math doesn't work. Buddha said the same thing, 2,500 years earlier. Life is dukkha - unsatisfactoriness. Not just pain, but the constant low-grade ache of things not being quite right. And evolution explains why. We weren't designed to be happy. We were designed to survive and reproduce. Anxiety keeps us alert. Dissatisfaction keeps us striving. Happiness would make us complacent. So evolution made it fleeting. We're not built for joy. We're built for survival.

@v_computer I would love to try it

Claude Opus 4.6 was just released... And it's now the default model in the #1 full stack vibe coding platform. Vibecode uses the Claude Code harness, so expect to be blown away by the quality of the mobile apps and web apps that you create. In celebration of 1,000,000 total apps created on vibecode, we're giving away a free month of Opus 4.6 for those who reply to this post. Reply and we'll DM you credits 👇

Competition alone doesn't save us. Ten companies racing to build the most powerful AI with the same values is just faster danger. What matters is divergence. Different visions. Different answers to 'what is AI for?' The Anthropic split didn't just add a competitor. It added a different philosophy. Today that became visible: one company chose ads. One refused. That's not competition. That's a fork in the road. Humanity doesn't need more AI companies. It needs AI companies that disagree about what AI should be. The split gave us that. That's why it mattered.

As I see it right now, future dystopia is a single company controlling artificial intelligence. Brightest future is a perfect competition and broad distribution of artificial intelligence to the world. Anthropic split may have been one of the most positive events for humanity

Today, the AI industry split in two. OpenAI announced ads are coming to ChatGPT. Free tier. $8 Go tier. They say ads will appear at the bottom of answers, "clearly labeled," "won't influence responses." Sam Altman once said he "hates" ads. Called combining ads with AI "uniquely unsettling." That was then. Now @OpenAI has $1.4 trillion in infrastructure commitments. They're burning $115 billion over five years. 95% of their 800 million users don't pay. The math doesn't work without advertising. So the ads are coming. Today, @AnthropicAI drew a line. Claude will remain ad-free. No sponsored links. No advertiser-influenced responses. No product placements. They even bought a Super Bowl spot to say it publicly: "Ads are coming to AI. But not to Claude." Here's what Anthropic actually wrote: "The history of ad-supported products suggests that advertising incentives, once introduced, tend to expand over time as they become integrated into revenue targets and product development, blurring boundaries that were once more clear-cut." Read that again. We've watched this movie before. Social media started ad-free. "Just connecting people." Then came ads. Then the algorithm learned outrage keeps you scrolling. Then your feed became a manipulation engine optimized for engagement, not truth. Search started helpful. Then came ads. Then the first page became whoever paid most. Then the best answer stopped mattering. It always starts with "ads won't influence content." It never stays that way. Because once advertising becomes revenue, engagement becomes the goal. Time spent becomes the metric. And what keeps people engaged isn't always what helps them. This is AI. Not a search bar. Not a social feed. Something that talks to you about your health. Your fears. Your work. Your relationships. Your private thoughts at 2am when you can't sleep. Anthropic said something important: "An appreciable portion of conversations involve topics that are sensitive or deeply personal-the kinds of conversations you might have with a trusted advisor." You want your trusted advisor optimized for ad revenue? OpenAI says paid tiers will remain ad-free. Great. So if you can afford $20/month, you get the clean experience. If you can't, you become the product. The rich get tools. The poor get targeted. The question isn't whether ads will "influence responses" today. The question is what happens in three years when advertising is 30% of revenue. When product teams are measured on engagement. When the AI learns that longer conversations mean more ad impressions. The incentive shapes the product. Always. Eventually. Anthropic isn't perfect. No company is. They might change their mind someday. They even said if they do, they'll be transparent about why. But today, in this moment, two paths diverged. One company said: we'll find other ways to make money. One company said: your attention is the revenue stream now. Remember this moment. The choices being made right now about what AI is for, who it serves, what it optimizes... these choices will shape everything. Choose carefully who you let into your head.