Hi, I’ve been constructing my median world, and I think I’ve used some ideas for incentive optimization and solutions that are rarely discussed here. Some parts (in my opinion) are novel, some are just value-dependent details and aesthetics. Some could actually be implemented (I haven’t heard of anyone treating ETF investing as a moral act, and mnemonics are indeed underrated), some are speculative. I thought about it mostly for my own amusement, but it has accumulated enough details that I thought it’s worth sharing, so feel free to criticize, my world is flexible and you can change it.

 

A Fraction of Global Capital is indeed the Best Currency

I tried to describe this in my previous post, but did it poorly, now I’ve found some new advantages.

Firstly, under the Efficient Market Hypothesis, total market ETF is the null hypothesis of investing. If you don’t have any unique knowledge, the only way to avoid dumber or smarter than the market is to buy the whole thing evenly. Any transaction means you value something above average market growth, thus adding your unique knowledge into the market’s collective brain. It is Goodhart-resistant, since nobody is printing money, nobody is wasting computational resources on useless crypto proof-of-work, nobody is overmining gold just because it is “valuable by definition”. For an ordinary mortal, there’s no need to think about “how to invest wisely” given this baseline option.

Secondly, what I missed last time is that the ETF becomes a benevolent world corpocracy, selfishly interested in altruistic cooperative global GDP growth. According to the Universal Ownership Theory, the ETF internalizes externalities. Growth of one company may benefit it locally but hurt the rest of the market. However, since the ETF owns the same share of all companies, one’s externalities hurt its whole portfolio. It has no favorites, it prefers general growth. If the fund owns 30% of all corporations, it can ensure 30% of votes for ESG across the board, so that abandoning aggressive ad strategies, voting against mergers, or implementing industry-wide standards via assurance contracts now requires just 20% allies, not 50%.

Fractal Dunbarocratic Equity Compensation

A CEO receives investment in real money - in ETFs - and pays department directors with company stock proportional to their best estimate of effectiveness. The number of directors is kept small enough that each one’s contribution can move the stock price, giving them non-symbolic motivation to work for company growth, which automatically grows their payoff, unlike giving Tesla stock to a worker who has zero impact and just treats it as cash at the market price of the stock. They also know each other personally, aligning them toward synergy instead of bureaucratic KPI compliance.

A director, in turn, can issue “shares of their department”, backed by the company stock they receive from the CEO, distributing them to their own direct subordinates. Again, the team is small enough that each contribution matters and people know each other. This continues fractally down to, say, project-level, where a sprint team gets project-shares, bought out at market price when the sprint ends. In this world, “capital/means of production belong to the workers” to a degree sufficient to motivate productivity. Owning shares of a particular company makes sense only if you invest your unique knowledge in it - otherwise you’d rather just hold shares of the entire market.

 

Geotax

I don’t like classic LVT, since land auctions capture man-made effects like “land is expensive because it’s in the city center”, which they shouldn’t. Instead, I’d tax precisely defined resources: soil chemistry and fertility, water, insolation, extracted iron tonnage, occupied water routes, CO2 emissions, plastic as an externality, etc. The tax just goes to UBI.

The tax formula is “I vote for a high tax on what I personally use less of”. At macro scale, this tends to reflect consumption inequality (planes are used more by the rich, so the poor vote for high air route taxes and get the consumption gap back as UBI) and demand elasticity (if you tax water too high, someone invents ultrasonic showers). Taxes can be several times higher than the market price, with the assumption that poor people will more than recoup it through UBI. Bureaucracy is simpler for people because firms pay the tax, consumers only see it indirectly in prices, except when voting.

Intellectual Property via Clean-Room Reverse Engineering

Imagine an uninhabited island. Nobody created it, so it belongs to all equally. If you pick a fruit, you deprive others of it, so you must pay a geotax. But if the island is a metaphor for a theorem or other knowledge - not created, only discovered - then the fruits are infinite or easily copied, so no payment is required, since you don’t deprive others by consuming knowledge.

But since the island is uninhabited, nobody can eat the fruit. The one who builds a bridge there can charge for passage - like the first discoverer of some knowledge. There’s no fixed monopoly term, because if you overcharge, others can build a second bridge and break your monopoly. If another team, in causally isolated conditions, solves the same problem independently, they get equal rights to charge for their own bridge.

But nobody wants to waste resources duplicating research. To prevent zero-sum games, prediction markets estimate the expected time/price until someone else would rediscover it, and compensate you for the head start your discovery provided relative to the counterfactual world. For this to work, there must be causally isolated “Clean Cities” that deliberately lag behind in information and are tasked with reverse-engineering monopoly patents, to ground the markets and resolve disputes.

Some knowledge (like antibodies found by auto-labs) has statistical discovery chances. Some requires random flashes of insight. Some inventions are “in the air” (multiple inventors of the telephone in one year). Others are ahead of their time. All this affects price. Creativity, like writing, may have effectively unlimited ownership term - no one else can write literally the same book, each bridge leads to a unique island. Inspiration, however, remains free.

Law

Simply electronic liquid democracy. I don’t see how complex quadratic/Dunbarocratic systems with unequal vote-price can be stable if votes can be arbitraged. The system can culturally drift toward Dunbarocracy (knowing personally whom you delegate to), but it is not required. You only vote for a “federal definition of non-aggression”, minimizing zero-sum conflict, tied to your desired compensation for risks.

Votes can legally be sold. But if you take money from thieves’ lobbies, you reduce your compensation in case of robbery. Risks are aggregated by insurers and collectors, financially motivated to reflect values honestly. Usually people just subscribe to recognized formulas of expected costs/risks, just as they do with optimal geotax votes, and use Experience Sampling Method to assess more subjective risks (in addition to all the other psychological benefits of this method).

Rights defense is usually free. Aggression gives the victim a metaphysical right to compensation, which they can sell to collectors. Private competing collectors exist, there is no monopoly on violence. Yet there is a federal monopoly on sanctioning legal violence, so an armed rebellion is illegal in the eyes of all other forces, who are eager to collect compensation. No punishments except financial, isolation of criminals happens voluntarily at the level of contract cities. Justice doesn’t aim for punishment or rehabilitation.
 

Housing (Here begins the value-dependent part)

My world doesn’t understand how you can build cheap modular prefab homes and not end up with an arcology. “City” and “building” are usually synonyms - why even go outside? This creates five-minute megacities using a few levels of capsule lifts/trains, simplifying logistics and opening new markets.

Shops as warehouses and pick-up points, where you physically carry stuff home, become archaic. Everything is home-delivered, with delivery costs negligible. For hedonic shopping, you have showrooms, tastings, fitting rooms - but that’s a tiny fraction of market. Cooking, laundry, cleaning, garbage disposal, storage - all centralized. The difference between a call and an in-person meeting inside a city nearly disappears. Thermal insulation gets easier, since city volume grows faster than surface area. And also…

Thin Clients vs. Personal Computers

Personal computers are legal, as well as physical removal of anyone who owns them into crypto-anarchist communes. Being allowed into civilized cities with your own computer requires a license and psychiatric check, much like firearm ownership.

Historically, dense housing made it cheaper to lay cables from a city server into every apartment, so the mainframe-terminal model never vanished. In this world it doesn’t carry the stigma of tyranny. It’s easier to record video of the entire server-boot process - from hardware manufacturing checks to launching formally verified post-quantum code, with transparent security audits under the rule of “everything known except the key”, data warehouses that only log and require physical access to breach - than to ensure nobody screenshots an NFT ape.

All code is open-but-unstealable, with per-file access settings, user experience like “as if you downloaded the Internet”, no difference between file paths and URLs. Function/library calls are billed per use, incentivizing clean documentation and APIs, destroying monolithic projects. Even forks pay royalties upstream. No thousands clones of the same calculator - just one ideal solution. No compatibility hell, since everything runs on the perfect hardware for it. Devices have no technical requirements beyond “show any streamed video”, while you rent exactly as much compute and memory as needed. Most problems are already solved, so zero-code builders cover everyday use cases.

Programmer effort shifts into 3D - making decent VR/AR interfaces. Without local compute, thin-client VR/AR becomes light, cheap and profitable. Centralization also removes file duplication on end-user devices, reduces downtime, improves cooling, and allows better hardware specialization (there is a difference between putting a rare-use chip in every PC vs having some in a city server farm). But these gains are minor compared to protecting intellectual property and making copying impossible when you literally have nowhere else to run it.

You can imagine how useful this is for AI risk management.

Energy

In hindsight, CSP (concentrated solar power) was obvious, and could have existed at least at proof-of-concept level since antiquity. Mythical Archimedes mirrors + Heron’s steam engine could have been combined theoretically, and over history, CSP was optimized enough to compete with coal during the industrial revolution. Scaling manufacturing just made it cheaper. They don’t increase efficiency much beyond today’s levels, but they’re reliable, so capital expenses with low opex pay off, since CSP plants can just work for 60+ years, unlike depleting PV. In colder regions, deregulated nuclear power is fine.
 

Ecosystem

Effective Altruists of my world sterilized almost all animals with gene drives so they can’t feel pain. Nature is radically simplified, mostly human-useful plants with a few symbionts (worms, stingless bees, etc.). Reserves maintain the last wild specimens, in case their populations ever need restoring.

Domesticated animals breed under control but can graze semi-wild. Forests are seen as pointless sunlight competition, and therefore they are mostly replaced with shrubs bred for ultra-deep roots to stop erosion and reach groundwater.

Without existing evolutionary templates, it’s hard for multicellular predators/parasites to reappear. Bio-risks remain mostly in diseases, but those are minimized too. So humans can live in nature comfortably.

Heredity Optimization

There’s a simple eugenics method to raise IQ by at least whole SD in the first generation, which is to stop caring about biological kinship. People already raise adopted children, and even children from affairs, so it’s not that deep of a value. If what you want is parental experience, social bonding, memetic rather than genetic reproduction, then a social couple can just pick which biological parents to have kids from.

They can buy sperm and eggs from, say, Einstein and Curie - gametes priced dynamically to prevent homogeneity - which already ensures top-0.1% selection each generation for any trait (face symmetry, happiness, whatever). These are fully natural children, no edits, just IVF and gamete freezing - which existed even before CRISPR.

Childbearing can be delegated to surrogates, which were unpopular until the new generations stopped feeling pain.

Absence of Pain

New generations don’t feel pain. Editing SCN9A-CIP spread quickly once possible. This world is already safe enough that the lack of nociception isn’t fatal - ergonomic measures like springy floors, with hot cooking/sharp tools left to professionals, everything under cameras with fast emergency response, sterile environment, etc.

Kids must wear caps/helmets, plus health checkups - but the subjective benefit of never experiencing pain outweighs it. Loss of smell can be fixed later with local gene edits after birth, but even if not, it’s still acceptable. Other edits are rare - they prefer well-tested monogenic changes or selection. But few modifications are as valuable as removing pain (though disabling myostatin would also be nice).

(In my median world, there are no Keepers talk-controlling all depressive negative utilitarians into voluntary genocide cryosleep under threat of isekai-immortality theory, so we had to satisfy those needs in other ways)
 

Beauty

No pain changes attitudes toward the body. Risk of injury rises, but medical compliance soars - kids raised with discipline, no need for anesthesia in surgery. And it enables soft-transhuman tweaks, like:

Total hair removal, head also, replaced by realistic wigs for easy style changes, easy-care with centralized special "laundries".

Nail removal, replaced by socket-implants with snap-on tips, instead of endless manicures.

Teeth removal, replaced by six-for-all equivalents - ideal smile, removable if needed (soft options to avoid biting yourself without pain perception).

More standard mods: skin polishing, botox, lipofilling, plus myostatin-off hypertrophy, plus gamete auctions including beauty traits.

Preference leans toward capex over opex. If one surgery replaces weekly routines, do the one surgery.
 

Internet Culture

The internet is strictly divided into personal and impersonal. Personal posts are visible only to friends or friends-of-friends. You can “keep a personal diary” with news about your life and your thoughts, and only those who actually care about you will respond. This removes the need for statistics, likes, or public validation.

The impersonal internet runs Kialo-like debates and builds complex wikis where data and arguments matter more than the pseudonymized person behind them. A hybrid of expert and blogger is generally seen as a failed specialization. Scripts written by experts can be presented to the public by arbitrary talking heads. This keeps parasocial dynamics, social comparison, and other harmful effects of social networks to a minimum - not fully eliminated, but reduced enough to lower net harm.
 

Religion/Philosophy

The local version of Buddhism says that since you had no control over who you were born as, you also have no control over what body you’ll wake up in tomorrow or at what moment you’ll find yourself. There is no “your suffering” or “your happiness” - only global metrics of suffering and happiness, and your expected utility of finding yourself as a random moment. The ideal is a total hedonistic utilitarian bodhisattva, absolutely coherent in treating the happiness of all beings as their own.

Reincarnation isn’t interpreted spiritually, but as the mathematical equivalence - there’s no difference between “you live all lives of all who have ever existed or will exist, over and over again” and “you’re frozen in a random instant in an eternalist universe where time is illusory” from the standpoint of mathematical expectation.

Karma is read as the fact that by harming another, you harm yourself, because you can’t guarantee you won’t be your own victim. By helping another, you yourself will experience the result of that help, “retrospectively maximizing your odds of finding yourself as a happy moment”.

Some can’t let go of the illusion of time, continuing to live as if their ego were real and tomorrow they will wake up as themselves, consuming benefits as if they’re theirs alone and not shared with eternity - much like any other weak-believers.

Earlier, such people faced social pressure toward more altruism, though their “sinfulness” was seen less as moral failing than as irrational refusal to pursue goals that actually benefit them. In the present of my median world, though, it’s universally accepted that the world doesn’t need much saving, altruism has no low-hanging fruit that demands panicked mobilization, and so this metaphysical current has de-radicalized, allowing illusions if they’re pleasant.

Not everyone shares these metaphysical foundations. That’s fine too. Unfortunately, my median world still needs religion as a stimulus for charity. It’s prone to trust-sincerity-cooperation, has nearly eradicated psychopathy, but remains weak in empathy and altruism. And even so, most charity is not given during life but through wills. And even then, many people manage to burn through their wealth exactly before agreeing to assisted suicide, leaving nothing behind.

Cryonics isn’t that popular. Many don’t want immortality, they want their body matter to be eaten and turned into a hedonium shockwave. But people still sign up, since it’s not that expensive and there are possible scenarios of resurrection before singularity, or of their brains being useful as information storage, or of their values being incoherent and immortality being actually good for them. Some aren’t convinced, and at least 10–20% just die their True Deaths.

Education

Visual mnemonics can be universal rather than “creative”, and are absurdly underrated. They’re even selfishly useful and require no coordination, so I’m surprised they haven’t been developed, and that the efficient market of our world somehow missed this. It’s a completely exploitable equilibrium.

Education here is mainly about helping people memorize exact data, breaking it into frequent elements as image-codes, with tons of examples for maximally algorithmized reproduction, plus mastery learning through tests that require high precision to pass.

It’s hard to combine this with Socratic education, since Socratics are too chaotic. So I treat them as just two different points on the Pareto frontier, optimizing for different things, maybe depending on subjective values. My median world isn’t that curious and doesn’t see much difference between discovering something yourself and consuming information, but Clean City welcomes all curious ones.