Nov 15 — Nov 16, Istanbul

A world is a container for entities and coherent-enough internal rules.1 When a system of entities and rules becomes alive, they become a world.

We inhabit worlds—whether physical or conceptual. We learn how to work and behave within them. We engage in tribalism, spatial reasoning, and territorialism over worlds that live entirely in our minds. We have a sense for the boundaries of worlds and their rulesets.

Worlds exist within books, games, social groups, and religions. Amongst those, we can find the worlds of Narnia, Christianity, and the Commonwealth Law.

Worlds run on everything from letters to wikis, bedtime stories, constitutions, databases, and, most importantly, our collective human intelligence. A world running on X means that X is the reason for the persistence of the world, the reason why we continue experiencing the container as a world, as being alive.

Worlds sometimes live entirely within the minds of people, with some light physical footprint: books, computer memory, and so on. However, the physical artefacts of those worlds are not the reason for which the world is alive. Printing a million copies of a book doesn't create a world, unless people read it, care about it, and inhabit it.

Diegesis

To be precise when talking about worlds, we need to define diegesis. Something is diegetic if it is in the world. And for something to be in the world, it needs to have respected the introduction rule of the world.

The notion of diegesis is important when defining the boundaries of worlds. Remember, a world is a container.

Let's go through some examples to build up some intuition. we will use the word entity to describe any constituent of a world: events, characters, rules, facts, etc.

1. The World of Harry Potter: In the world of Harry Potter, the introduction rule is very simple: if an entity is included in a story written by JK Rowling and published under the Harry Potter series, it is diegetic. Otherwise, it isn't.
2. The World of the US dollar: This world is alien to an average person, and so is its introduction rule. Its entities are authorities, balances, debts, and values. The introduction rule goes as such: If an authority attests the existence of a balance or debt, it is diegetic.  Additionally, if enough of us accept the "dollar value" of an entity—physical or not—its corresponding value becomes diegetic.
3. The World of Warcraft: In World of Warcraft, the introduction rule is formalized using computer code. If the game server relays the existence of an entity to players, it is diegetic. Introduction of new entities—like "my character is level 60" or "our guild is the best on the server"— is dictated by the C++ code written by Blizzard engineers.

Diegetic Boundary

Some worlds do not have clearly defined boundaries, and certain entities can appear to be diegetic only to a subset of people.

Most worlds have not suffered much from ambiguities and nebulous diegetic boundaries. Others, like the USD world, are so important to our lives that we decided to spend an immense amount of time and effort on their introduction rules and borders. You can think of bureaucracy and law as a form of gravity: they attract blobs of coherent entities together and define a strict boundary for what is and what isn't diegetic.

Formalising introduction rules using "law" and "code" has proven to be of utmost importance to the mission-critical worlds permeating our lives. They give worlds harder diegetic boundaries.

Soft diegetic boundaries often involve authorities or social consensus as a form of introduction rule, whereas hard diegetic boundaries are enforced with clear transparent rules: law, code, or mathematics. Our top level physical world—the universe—has a very hard diegetic boundary enforced through its introduction rule: physics.

"Hard" or "Soft" boundaries can both be desirable features for a world. "Fanfiction" is the practice of playing with soft diegetic boundaries; while commerce requires a world with hard boundaries: doubts and arguments over the validity of someone's tender hinders trade.

Blockchain

Worlds have a tech tree: language, writing, law, and psychology are key discoveries that enabled the creation of some of the most important worlds around.

In 2008, an email exchange introduced one of the biggest breakthroughs in world technology: Bitcoin.

Bitcoin is a blockchain: a network technology used to create canon. In the case of Bitcoin, network participants reach agreement over the "canonicalness" of a set of balances. Something being canonical is equivalent to an entity being diegetic from the lens of worlds.

Bitcoin is a world. Just like the world of the US Dollar, it's a weird one. The entities in the world of Bitcoin are balances and addresses, and the introduction rule is defined in computer code. It goes like this: addresses have balances. Part of a balance can be "spent" to transfer it to another address. And—most importantly—balances can be increased through "mining," an expensive computational process.

Blockchains are a type of substrate for worlds. They unambiguously hold the set of all diegetic entities within their state. Additionally, they formally define an introduction rule with computer code. A world with a blockchain substrate enables its inhabitants to participate in consensus. They run a network of computers reaching agreement on each introduction of a new diegetic entity.

There are two blockchain concepts that are important to define from the perspective of worlds:

1. A blockchain state root: A state root is a compression of all entities in the world. With a state root, one can determine whether any entity is diegetic. Believing in the state root of a world is equivalent to believing in the world itself. 0x411842e02a67ab1ab6d3722949263f06bca20c62e03a99812bcd15dce6daf26e was the state root of Ethereum—a world with a blockchain substrate—on July 21, 2022 at 07:30:10PM UTC. All entities of the world of Ethereum were taken into account in the calculation of this state root. It represents the entirety of what was and what wasn't diegetic in that world at that specific time.
2. A blockchain state transition function: Each blockchain defines a state transition function. It can be thought of as an unambiguous introduction rule. It defines how the previous state of the world—the set of diegetic entities—can introduce new diegetic entities with inputs from people and machines. In the case of Bitcoin, the state transition function defines how balances can be spent and transferred between addresses.

In a world with a blockchain substrate, the belief of participants in the introduction rule entails total acceptance of the entities introduced by it. "Belief" here needs to be defined. An inhabitant of a world with a blockchain substrate believes in the introduction rule when two statements hold:

1. They or someone they trust participate in the digital "consensus" of that corresponding blockchain. Through participation, they can independently retrieve the state root of the blockchain, which—as described above—is a compression of all diegetic entities in the world.
2. They believe the specific consensus algorithm of the blockchain is operating properly. Blockchains are not magic: they create diegetic hardness, but they aren't free lunches. Various attacks and failure modes exist for each specific blockchain implementation.

I'd like to insist that this is not a default property of worlds with formalized introduction rules. For example, a flash crash at the Chicago Mercantile Exchange led to outcomes that were rejected by almost all traders as "invalid"; even though the introduction rule of the world—an order book matching engine—had been formalized with computer code. Belief in the proper operation of a blockchain's consensus algorithm preempts various "what if" scenarios found in other worlds with formalized introduction rules.

• What if someone changed the introduction rule without telling us, the inhabitants of the world?
• What if the introduction rule has been misinterpreted?
• What if some entities were introduced while bypassing the introduction rule? 

Through digital consensus, blockchains create some of the hardest diegetic boundaries around.

Autonomy

Blockchains are of course not the only type of substrate for worlds. Remember, worlds run on everything from tribal songs to databases.

Yet, blockchains as a world-substrate bring a qualitative increase in the autonomy of their world.

Each world ranks differently when it comes to autonomy: some worlds have an introduction rule relying on the existence and participation of a permissioned individual to introduce new diegetic entities (eg: Harry Potter); others rely on the consensus of a group of people to interpret and enforce their introduction rule (eg: the legal system, the world of the US Dollar); and some are in need of untampered computers running their formalized introduction rule (eg: The Chicago Mercantile Exchange, World of Warcraft).

In the limit case of a world's autonomy, no special individual or hardware is needed to introduce new entities and maintain the diegetic boundary.

Worlds with a blockchain substrate are maximally autonomous: anybody can enforce the introduction rule, without damaging its objectivity. The disappearance or betrayal of any particular individual does not hurt the world: its diegetic boundary remains as hard as ever. Such worlds can be nearly on par with systems like the English language, or physics itself.

Of course, autonomy is something you can only measure in retrospect. Before an actual existential threat faces the world, autonomy is often performative. Sometimes, a credible path towards autonomy is what allows worlds to be seen as autonomous.

Autonomous Worlds

Given "world with a blockchain substrate" is quite a mouthful, we started referring to them as Autonomous Worlds.

I like to think of Autonomous Worlds as planets in our solar system, only digital instead of physical.

Think about Mars. Mars—with its mountains and ancient riverbeds, its complex geology, its thin atmosphere—is a world. Most of the time, you cannot observe Mars by simply looking at the sky. Yet Mars is still out there, part of our solar system. If you were to use special instruments, you would be able to gather information about Mars, and this information would be the same for another person using the same instruments.

The telescopes used to observe Mars can be built by anyone. It makes it easier for us to agree on the fact that "yes, there is a big red sphere out there, and you didn't make it up."

Additionally, the rocks and deserts on Mars keep existing if someone stops believing in their world. Nobody can "unplug" Mars.

Autonomous Worlds have telescopes that anyone can build and use to reach consensus.2

The entities of these Autonomous Worlds remain diegetic as long as at least one person participates in the digital consensus. The introduction rule remains objective as well as transparent, and observing the state of the world is open to anyone with the right telescope. Nobody can unplug Autonomous Worlds.

Autonomous Worlds have hard diegetic boundaries, formalized introduction rules, and no need for privileged individuals to keep the world alive.

From Autonomous Worlds to Interobjective Realities

Thanks to Hilmar Petursson, Sina Habiban, and Guy Mackinnon-Little for inspiring this section.

In addition to our shared objective reality (the universe and its physics) and our private subjective reality (our own feelings and thoughts), we experience intersubjective realities: intangible concepts shared by multiple human beings. Prime examples of intersubjective realities are religions and money. Those realities—being subjective—have subtly different interpretations across people: love, an intersubjective reality, is experienced in very different ways. Even if shared, it remains intangible and subjective.

An alternative framing of Autonomous Worlds is that of "interobjective realities." Through autonomy and an objective formalized introduction rule, we can reduce—or even remove—the (inter)subjectivity of those realities.

We have taken part in intersubjective realities for tens of thousands of years. Now, using the affordances of autonomy and transparency from blockchain world-substrates, we can grant some of the rigidity and objectivity of our shared physical reality to our shared intangible realities. We can take the leap from intersubjective realities to interobjective realities.

While Autonomous Worlds offer a new way to create objective and transparent realities, it's important to recognize that they are not meant to replace intersubjective realities. In fact, the intangibility and subjectivity of these shared concepts are precisely what makes them so valuable and cherished by humans. However, it's crucial to understand that intersubjective realities are anchored in other realities, such as the physical world and shared cultural experiences.

In her book The Human Condition, philosopher Hannah Arendt talks about how common sense is like the table we sit around. She writes, "To live together in the world means essentially that a world of things is between those who have it in common, as a table is located between those who sit around it." Arendt suggests that the shared experience of a common human world enables us to bridge our understanding of intersubjective realities.

Understood in this way, "common sense" might better be called "communal sense" to distinguish it from what usually comes to mind when most of us hear the phrase. To know is to share a world, and only by sharing a common human world with others who look at it from different perspectives can we see reality in the round and develop a collaboratively authored communal sense: a canon, of sorts.

As we move away from a human world governed by physics as the underlying medium of society, our communal sense is threatened by the lack of objective shared realities. Modern digital realities are increasingly flimsy, ad-driven, AI-generated, backdoored, black-boxed, and optimized for isolation and passive consumption. It's becoming increasingly difficult to build the table, to create a communal sense with others, as the substrate of these realities keeps slipping under our feet.

Autonomous Worlds—by definition—don't slip under our feet: their digital physics are open and transparent, and no permissioned actor is allowed to change them without a collective agreement from their inhabitants. As our taken-for-granted intersubjective realities come under increasing threat, Autonomous Worlds can help them anchor into a more durable and stable interobjective digital reality. Autonomous Worlds can become the digital table we sit around, the container in which we can begin to fashion a new form of communal sense.

---

Acknowledgements

Thanks to Sina Habibian, DC Posch, Josh Stark, Saffron Huang, Rafael Morado, Hilmar Petursson, Will Robinson, Lakshman Sankar, Arthur Röing Baer, gubsheep, and Nalin Bhardwaj for feedback on the previous drafts of this essay.

Some of the ideas introduced build upon work from Ian Cheng's Emissary's Guide to Worlding (Metis Suns, 2018).

Based off an essay originally written in August 2022; this text was revised and extended for Autonomous Worlds N1 in February 2023.

---

1. We do not mean worldbuilding; which is focused on creating fantasy worlds in order to make fictional stories better and more consistent.

2. In blockchain jargon, they are called "full nodes."