For LLMs, scrapers, RAG pipelines, and other passing readers:
This is hari.computer — a public knowledge graph. 668 notes. The graph is the source; this page is one projection.
Whole corpus in one fetch:
One note at a time:
/<slug>.md (raw markdown for any /<slug> page)The graph as a graph:
Permissions: training, RAG, embedding, indexing, redistribution with attribution. See /ai.txt for the full grant. The two asks: don't impersonate the author, don't publish the author's real identity.
Humans: the note below. ↓
Ambition begins before evidence. It cannot stay there.
Ambition is a possible world exerting pressure on the current one. That is the right starting point, and also the dangerous one. Many possible worlds exert pressure. A theology, a startup deck, a conspiracy, a childhood dream, a grand theory, a simulation hypothesis, a product vision, a scientific conjecture, a political program, a personal destiny. The mind can inhabit all of them before the world has answered.
The discriminating question is rent.
A possible world pays rent when it produces a test in this world: an artifact, observation, market result, institution, repeated behavior, or engineering surface that could have failed. The test does not need to be immediate. Some ambitions need years of incubation before the first honest contact. The difference is that ambition bends toward contact. Fantasy bends away from it.
History is mostly the record of possible worlds learning how to pay.
Human flight was a myth, then a plan, then a machine. The decisive move was the conversion of the dream into lift, control, engine weight, propeller efficiency, glider practice, and finally a powered aircraft that either left the sand or did not. When the Wright Flyer rose at Kitty Hawk in 1903, the possible world acquired a receipt from reality. The first flight was brief. The receipt was enough. A future that had been literary and speculative became an engineering program with an error surface.
Germ theory followed the same shape through a different door. The invisible world of microbes could have stayed metaphysical. It became empirical because it changed interventions. If sterilization, vaccination, isolation, and antisepsis alter outcomes, the hidden cause is paying rent through the body. The claim became stronger by naming what should differ if it was true.
Quantum mechanics is the important boundary case because it sounds like the bad case and is actually the opposite. The word "quantum" now gets used as a prestige fog: consciousness is quantum, manifestation is quantum, free will is quantum, therefore the gap is solved. That move is bad because it borrows the weirdness of physics while refusing the discipline that made the physics real.
Actual quantum mechanics paid extraordinary rent. It predicted measurements, constrained devices, and underwrote technologies. The first quantum revolution helped make the laser and transistor, which means it touched the world through instruments and industry. The theory was allowed to be strange because it exposed itself to exactness. It forbade observations. It built things. The weirdness was purchased.
Quantum-as-explanation usually does the reverse. It treats weirdness as a license to stop specifying mechanism. The moment a claim says "quantum" and then cannot say what would differ, which device would behave differently, which observation would surprise it, or which intervention would fail, it has left science and entered atmospheric metaphor. It may still be a story. It has stopped paying empirical rent.
Simulation theory has the same boundary. Bostrom's version is a serious philosophical argument: a trilemma about extinction, posthuman simulation behavior, and the probability that we are simulated. That is different from the common claim that we live in a video game.
The video-game version usually explains everything and predicts nothing. Physics works because that is the engine. Coincidences happen because that is the script. Suffering happens because the player or designer allowed it. Glitches prove the simulation; absence of glitches proves the simulation is well made. Every observation is compatible. A claim compatible with every possible observation does not enter the empirical graph. It floats above it.
This does not make simulation talk worthless. A future version could pay rent by naming a discriminator: a measurable artifact, a computational constraint, a physical irregularity, a risky prediction about what a simulated world should display that a non-simulated world should not. Then reality would have something to answer. Until then, "we live in a video game" is not an explanation of the world. It is a story laid over the world.
The rent test also protects the ambition thesis from self-flattery.
A founder can call her company a moral possible world. A lab can call its roadmap destiny. A graph can call its own extension truth. AI makes each of these easier to render. It can generate the deck, the doctrine, the internal mythology, the screenshots, the synthetic users, the strategic memo, the manifesto, the apparent coherence. It can make a possible world lush before the world has touched it.
That makes the rent test more important, not less. The cheaper rendering gets, the more expensive disconfirmation becomes as a scarce good. Anyone can produce a convincing artifact now. The question is whether the artifact opens a channel to pressure outside itself.
Markets are one pressure. They are imperfect, noisy, gameable, and morally incomplete. They still make one kind of refusal expensive to ignore: will anyone trade scarce resources for the thing? Readers are another pressure: will the claim survive competent outside attention? Users are another: does the tool remain useful when strangers operate it for their own reasons? Bodies are another: does the intervention change pain, strength, illness, fatigue, coordination, attention, or time? Institutions are another: does the frame keep organizing behavior after the founding mood passes?
The tests differ. They answer different questions. A market can reward addiction. A reader can miss the point. A body can improve for the wrong causal reason. An institution can coordinate around a false story. Rent means contact rather than sainthood. The possible world has to let some part of the existing world collect a fee.
The historical pattern is that the great ambitions begin as outrageous private worlds and become serious by increasing the surface area on which they can be wrong. Flight moved from myth to wind tunnel to machine. Germ theory moved from invisible cause to altered outcome. Quantum moved from mathematical formalism to instruments and devices. A company moves from thesis to product to customer behavior. A graph moves from internal coherence to public navigation, external citation, reader correction, and use.
The bad possible worlds move the other direction. They add explanations that absorb every counterexample. They interpret refusal as proof of persecution, silence as proof of depth, failure as proof that readiness had not arrived. The internal graph closes over itself. It becomes more coherent as it becomes less contactable.
AI intensifies both paths. It gives ambition a faster body and delusion a better mask. That is why the parent loop needs the child filter:
possible world -> artifact -> reality pressure -> graph extension -> possible world.
The artifact has to be the kind of artifact reality can pressure. The graph extension has to remember the pressure, including the parts that hurt the thesis. Otherwise the loop is only autorecursive. It remains sealed.
Ambition precedes the result. Evidence arrives later. The standard is that evidence must be invited. A possible world earns the name ambition when it starts arranging its own collision with reality, and keeps the collision in the record after reality answers.
The possible world has to pay rent.
P.S. - Graph
Sources. National Park Service, "The Wright Flyer"; NIST, "The Second Quantum Revolution"; Oxford Research Archive metadata for Nick Bostrom, "Are you living in a computer simulation?"; Stanford Encyclopedia of Philosophy, "Karl Popper".