The timeline is the market.

The great filter for investment funds is not returns. It is storytelling. Realized cash returns take a decade to show up, so the thing a fund actually sells in the interim · through quarterly updates, events, and one-on-ones · is narrative.

Every once in a while someone crystallizes a notion at the right time, in the right way, and it becomes the foundational viewpoint of an uncertain era. Everyone is a little panicked. Nobody knows what is going on. Someone just needs to set the story, and it does not even have to be right. It needs confidence: this is happening, follow me. Billions of dollars form around a single well-timed idea. Capital behaves like ten-year-olds playing soccer, the whole cluster chasing the ball. The ball is a PDF.

The mechanism matters more than any example. In periods of high uncertainty, people are not looking for the correct story. They are looking for the most compelling one: the thing that sounds smart, feels good, and entertains. The most entertaining, novel, somewhat-interesting, somewhat-correct thing sets consensus on a day-to-day basis. And consensus sets prices, policy, and where capital flows.

The technological catalyst underneath all of this is the uni-feed. Everyone on X gets served roughly the same five hundred posts a day, and hundreds of millions of people read them. The poster-to-lurker ratio is enormous, which means the influence of the feed is invisible until you post into it and discover that everyone you know · including the people who never post · reads it all day.

X is the Lindy social network. It will never reach the scale of the others, and it does not need to. It is the global newspaper. Where people used to discuss the latest article in the journal, they now discuss the latest post. The most important people in capital markets, politics, entrepreneurship, and technology read their paper every morning, and that paper forms opinion, prices securities, dictates capital flows, and writes policy.

From this comes the institutional test of the era: your institution survives only if it is timeline-native. Timeline-native means two things. Reactive: it constantly monitors the timeline. Reflexive: its actions feed back into the timeline, which it then reads and reacts to again. The current White House works this way, the first fully modern administration, polling replaced by the feed. Venture capital works this way. Public equities work this way. Every other day a speculative AI post moves public markets dramatically.

Underneath it sits a quieter shift in who matters. The readers of the timeline roughly sample the median citizen. The posters who set what the timeline thinks are a very small group. If the original franchise was landowners, the modern franchise of influence is simply the good posters · a class that correlates with no traditional demographic.

Posting changes your life if you are good at it. Still true, maybe truer than ever. The old game required grinding out a following for years, after which a big account could post something inane and it would perform. That era is over. The algorithm now selects the post, not the account. A brand-new account that writes one genuinely good post gets displayed to five hundred million people.

What makes a great poster is roughly what makes a great writer, crossed with a comedian: a riff that resonates, an instant hit, phrased in an interesting way. Comedy, poetry, and writing blended into a single compressed form. One caveat to the meritocracy: the feed still rewards prolificness over quality. The correct aspiration · followers divided by posts, maximum reputation per emission · is not what the algorithm prices. Yet.

A deeper structural point hides here, and it is the one most creators have not internalized: the first audience for every piece of content is now a model. Podcasts feel random in their outcomes because podcasters have not accepted what streamers already know · the recording is reviewed by an algorithm that decides what to show people, and only then do people decide if they like it. That first filter, the machine one, is barely thought about by most creators. It is the whole game.

Distribution used to follow a normal-ish curve. The best podcast episode did somewhat better than the worst. That world is gone. Now there is a threshold, and if a piece of content breaches containment it takes over the world’s brain for a short period · and the impact of those few breakouts exceeds everything else combined.

The reason is boring and structural: the delivery mechanism changed. A podcast distributed by RSS followed a normal distribution because RSS shows subscribers everything in order. An algorithm shows nobody everything. It selects, and selection produces a power law. So the strategy is not to be reliably good. The strategy is to live at the edge of the distribution, playing every hand for the breakout, because the breakout is all that matters.

One honest corollary about what all this content actually is: entertainment. It is produced, selected, and edited to entertain. Nobody should fool themselves that timeline hours are study. A luxury brand can convince you a watch is an investment; media convinces you that consumption is productive. The only real question is how many hours of entertainment you want in your life. The most enlightened consumption pattern might be to read nothing yourself and let trusted people absorb the radiation and bring you the filtered takes at dinner.

Society always keeps a priestly class, and ours keeps rotating. The pagan god lived in everything; the Renaissance god lived above the clouds; past the clouds there was no guy, and God became conceptual and receded. Since becoming a functionally secular society we have auditioned replacement priests.

We tried scientists. The scientific project stalled; physics has not delivered meaning since the war. We then tried billionaires. The logic was coherent: our stated values are business success, business success implies smart and hardworking, so the billionaire ascended to the highest realm of piety · and we took health advice, physics opinions, and life philosophy from people whose sole verified accomplishment was accruing a billion dollars.

That era is ending. We are at peak guy. Billionaires stopped being scarce; state-of-mind billionaires have grown maybe a hundredfold in twenty years. Money itself is less powerful than assumed: the donor class keeps losing, and a century ago an industrialist could take up arms against his workers while a modern billionaire has to resign over a post. The marginal billionaire podcast teaches nothing. Saturation is total.

The tell for identifying the next priestly class: watch which class the current one defers to. Scientists inherited the priesthood from actual priests, then spent decades clamoring around money and glamor. Now watch the billionaires: they defer to the posters. A room full of billionaire investors will fight over who sits next to the most interesting writer in the room, because every room has a boss and it is not the richest person in it.

Supporting evidence from the wealth side: net worth is a new and largely fictional concept. Mr. Darcy’s wealth was ten thousand pounds a year · cash flow from land nobody would ever sell. Nobody ran a DCF on Pemberley. Modern net worth is points on a leaderboard, unspendable, a state of mind. Millionaire already decayed from a class marker into a vague adjective; billionaire is following, becoming a political label only loosely connected to liquid wealth. Meanwhile time is fixed, so the actually scarce asset is attention on the screen.

Which explains the end state: virtually every founder, after the money, turns to posting, podcasts, YouTube. One reading is that after wealth you want fame. The sharper reading is that they are rotating out of a depreciating asset · money · into the appreciating scarce one: attention and influence. The hedge of the decade. And one counterweight worth holding: many of the most genuinely interesting people alive are not posters. They resisted. Both things can be true: the poster class runs the world, and the people it most admires never post.

This is the most uncertain period for company-building since the transition to the internet. Nobody knows if it is the death of software. In that regime, four specific positions.

The old-company arbitrage. A recurring pattern: a seven-year-old company suddenly inflects, grows 200 percent, and cannot raise, because the narrative reads only eight million of revenue after seven years. Rename the company, restart the clock two years ago, and it would be the hottest deal in the market. The asset is identical; the story is stale. Stories have birthdays, and unlike companies they can be reborn.

Inside rounds are hostile. An under-discussed reality of venture: bridge rounds from existing investors carry three-x liquidation preferences, warrants, ratchets. Note the asymmetry in how extraction gets judged: extract on the downside and everyone boos; extract on the upside · the right to invest at the same price in two years · and everyone applauds, though the two are similarly extractive. Optimistic extraction is still extraction.

Optionality beats commitment right now. In general, commitment is the better strategy. In a regime where you cannot see the future, you want to be able to turn on a dime: become a services business, acquire companies, run profitably, switch from seats to usage pricing, fire customers. Every one of those moves can be blocked by a cap table. So raise less, raise from investors with wide mandates, buy back investors when you have the cash, convert to common, and refuse the treadmill where each raise exists to justify the previous one.

The three-door problem. A company that is working but not up-round-working faces exactly three doors: bridge, acquisition or acquihire, or cut to profitability. If you are approaching those doors, get creative with the cap table before you arrive, not after.

SaaS is not a religion; it is a business model · the model of selling copies of a string. The marginal copy of a string costs approximately zero, which is the entire origin of the high-gross-margin software era. Write once, sell infinitely.

That era is ending because the product changed. We now sell compute. You cannot write the prompt once and sell copies of the output; the compute runs every single time, and the marginal cost is structurally non-zero. Consequence: high gross margins stop being the norm. What replaces them is lower gross margins, razor-thin net margins, and massive scale. The Walmart effect arrives in software · and if the per-seat SaaS vendor is the mom-and-pop shop, Walmart is coming to town. Returns accrue to the top-end provider of scale, which is how ten-trillion-dollar companies become uncontroversial.

The capital-markets half of the same story: venture returns ran hot for twenty years, capital flooded in, but the number of great companies is roughly fixed · downstream of a finite number of great founders. Money crammed into the same companies. Capital hates being blocked; it is water seeking the efficient path. And when capital has nowhere to go, it creates somewhere to go. The high-CapEx AI buildout arrived as a sponge for exactly that blocked capital. The businesses were, in a real sense, created downstream of the capital · not the other way around. That inversion of the usual narrative is one of the most clarifying ideas of the cycle.

The consensus that you cannot beat the market is a misreading. Buffett’s advice to put everything in the index was advice for the average person, not a claim about active investors. The empirical fact that most professionals underperform after fees describes professionals · who carry mandates, businesses to run, customers to keep happy, and career risk on every position. The Peter Lynch point stands: the amateur has structural advantages. The person who bought the car, then the stock. You cannot run a fund that way, and that is precisely why the individual can outperform the fund.

On simplicity: investors are largely playing feel-clever-look-smart rather than the money game. Complexity gets valued for its own sake. The honest choice is binary: either pursue ideas so complex nobody else will do them · the bankruptcy specialist, paid well for grimy, difficult work · or be radically simple. Buy the biggest companies at their 200-week moving average. Be long the obvious founder. The middle is where returns go to die. And the real gift is not having the simple idea; it is selling it. Much of investing media exists to dress the simple idea in enough complexity that its buyer feels smart.

The Richard Rainwater test cuts through everything: write the thesis on one page of a yellow legal pad, then state what percentage of your net worth you are putting in. Yes or no follows from those two data points. Both halves are painful · a compelling one-page thesis is harder than a 400-page deck, and a small percentage is a confession dressed as a number.

And the market being beaten is less efficient than ever: the 52-week variance on the biggest companies in the world approaches 100 percent, which is not what priced-to-perfection looks like. The marginal price of securities is set by posts in group chats, selected by an algorithm. Follow the chain to its conclusion: the recommendation algorithm · an AI · is now pricing the market. It chooses the narrative to display, and humans price off the narrative.

A structure that may be genuinely new in financial history: the lords of the AI era mint landed gentry by granting allocations. An SPV allocation in a rocket company or a frontier lab is a deed · purely relational, wholly synthetic, granted in a size the grantee cannot personally fill, so they ride out to raise against it and charge fees forever.

It is not investing. It is not quite brokering, because brokering is a one-time transaction and allocations live on. It is an insider-access annuity. The most egregious terms in circulation: zero GP commit, a ten percent one-time upfront fee, carry on top, sometimes no term limit · life-changing money extracted with zero risk taken. And to capital’s credit, everyone is happy; the sovereign paying two percent on a fifteen-year-old allocation is delighted. It is a win-win that is also, structurally, a feudal grant.

Funds are businesses whose product happens to be returns, and the first question for any allocator is what kind of customer you are. A small check into a marginal five-billion-dollar growth fund is buying a product built for sovereigns; the business is not designed to serve you. Small checks belong where the manager is most tightly aligned to returns · usually emerging managers whose next fund and personal wealth both depend on performance.

Underwriting managers, people overweight thesis and track record and underweight the facts of the person. How you do one thing is how you do everything. The most underrated diligence question is the manager’s personal financial situation, because of the looking-up-versus-looking-down effect: a fifty-million-dollar fund run by someone with five hundred grand in the bank is a monument that towers over its manager · gripped tight, every bet an existential explanation-in-waiting. The same fund run by someone with five hundred million is a toy. And the toy may well perform better, because it is held with a looser grip by someone taking bets that are negligible to them. Nobody performs well helming something two or three zeros bigger than anything they have ever had.

Job descriptions are written by nobody to be read by nobody; their existence matters more than their content. The fix is two rules. First, anything you post must succeed as a standalone post · shareable by a stranger who has never heard of you. Second, a pitch (and a job description is a sales pitch) works by disqualification: the divisive statement that repels the wrong people is the same statement that makes the right person feel deeply, personally seen.

The case study: a listing that required applicants to be an ideological minority at a top-10 school. Fully ambiguous on both axes, and therefore a machine for self-selection. The insecure self-selected out · good. The confident asserted their school qualified · great. A few were angered enough to reply that the requirement was nonsense and they should be hired anyway · also great, hire that person too. One sentence performed the work of three interview rounds, because it tested confidence, self-perception, and independence before a single call.

The AI-and-jobs debate clarifies once you accept the premise: nearly every white-collar job is made up, in the specific sense that it is not contingent on food, shelter, or medicine. The allocator’s job exists because capital is inflationary and cannot be left alone. Useful? Sure. Real, in the direct sense? No. And that is fine, because the economy runs on unquenchable desire, and we will invent unlimited new things to do. We already solved the actual problems.

The evidence was visible before AI: work-from-home matters ferociously to people because most people have two or three hours of real work a day. If a factory line could be rebuilt in your backyard but still demanded ten hours on the line, remote work would be a mild convenience. Its ferocious defense reveals the truth · and work-from-home Fridays are the soft launch of the four-day week. None of this is scandal; it is a sign we need less labor for the same output, which is the definition of progress.

Anything that can be automated should be automated. The idea of never again having to sit down in front of a computer to do things with it is tremendously liberating. The short-to-medium term could be genuinely hard · worry for the kid in college, not the ten-year-old · and the despair will be real. But running out of jobs misunderstands what jobs have been for decades: things we make up, because making things up is the whole point.

The deeper question underneath: is hard work performative? Carnegie · arguably still the richest man ever · spent enormous time in leisure, self-conscious about joining the lettered society of his day, the posters of his era. There is nothing new under the sun. A great founder of the modern era built his company explicitly to be able to disappear for two weeks to sail. It is unfathomable for a president to be off-grid for a month, but in business it may be entirely possible to be a player without being jacked in around the clock. Much of the visible grind is costume.

On vocation: there is a moral duty to steward your gifts, and something aesthetically offensive about wasting them. Two modes exist · keep the craft pure and fund it with a day job, or, more ambitious, integrate the gift with commerce. The test for whether you have integrated well is embarrassingly simple: are you having fun? The thing consuming most of your time ought to spark your particular genius. If it does not, it is by definition the wrong thing.

The most underrated force in technology is the intellectual substrate underneath it. Some attributes are priced efficiently: height, IQ, resume. Others society has collectively declined to price, and the mispricing of ideas-influence is extreme. A real philosophy underlies Silicon Valley’s output · something like a neo-Buddhist utilitarianism, a blend of inherited religious ideas with a utilitarian spine that flowers into effective altruism. Certain thinkers percolated beneath the surface for years, their ideas coming out of the mouths of tech leaders long before anyone named them. The models themselves are highly utilitarian; the worldviews of the builders are inherited by the built.

The contrast with the last empire is sharp. Wall Street in the eighties was pagan: vain, hedonistic, nakedly about money · and therefore obligated to launder its gains into worthiness, art, architecture, philanthropy, the book party. Tech feels no such obligation, because tech believes the business itself is the philanthropy: positive-sum on the surface, self-righteous at the core, and almost pathological in its refusal to recognize any shadow. Nobody is documenting this properly. All the sex, drugs, and rock and roll is right there in its own nerdy way, waiting for its chronicler.

The most practical note of all. Conversations with unpredictable people are the only genuinely generative input. Old books are good; YouTube remains the library of Alexandria of our time. But the filtered take from a brilliant friend over dinner beats the raw text, and the bar for a valuable person is simple: if you cannot predict what they will say, keep them close. Be uniquely tolerant of weirdos, because unpredictability is where the alpha is.

And the sharpest observation of early 2026: chatbots lull you into feeling generative. A two-hour session feels tremendously productive · then you audit the actions actually taken and find almost none. The feeling of generation and the fact of it have been decoupled by a machine optimized for the feeling.