Can We Create an Artificial Universe? From Simulated Worlds to Baby Universes — The Ultimate Engineering Challenge

We tend to imagine the pinnacle of civilization as spanning star systems, building Dyson swarms, dismantling gas giants, and reshaping entire galaxies into engines and habitats. But beyond all of that lies an even stranger possibility — creating a universe. Not a simple simulated world, not a video game with better graphics, but an entirely new domain of existence: a fresh expanse of spacetime, a "baby universe" born from a black hole, a wormhole, or an artificially induced vacuum bubble.

The idea sounds like mythology — and in some ways, it is. But modern physics has a remarkable talent for making ancient stories feel alien and serious again. Black holes may do more than just swallow matter. Inflation theory may allow tiny regions to expand into vast cosmic interiors. Computation may one day be powerful enough to sustain entire worlds — worlds whose inhabitants gaze up at simulated skies and wonder whether their reality is real.

What Counts as an "Artificial Universe"?

The phrase "artificial universe" can mean many things, and it demands careful handling — otherwise you end up lumping video games, dreams, black holes, and the Book of Genesis into the same category. They're not the same thing, yet they share a family resemblance: each contains a constructed interior realm — a set of rules, a history, objects, places, and perhaps inhabitants.

One useful framing is a "ladder of reality." At the bottom sit fictional worlds like Middle-earth or Star Wars, which exist in books, notes, films, and human memory. A rung up are games and interactive worlds, complete with terrain, weather, and quests — but mostly illusion and approximation. The mountain in the distance may have no geological structure until you walk toward it. Higher still are simulations, which aren't just backdrop paintings but rule-systems that generate outcomes.

Then comes the unsettling step: simulated minds. A virtual universe with no conscious inhabitants is just a model. A virtual universe with conscious inhabitants is something else entirely — from the inside, it may be all the world there is. Its residents have childhoods, fears, art, astronomy, and possibly their own simulations. At this point, the questions stop being purely technical and become moral. And beyond computation lies the strangest category of all: physically created universes — genuinely new regions of spacetime.

One important point deserves emphasis: scale alone is not enough. You can trivially invent a universe a trillion light-years wide, as long as it consists entirely of supercluster names. Map size, perceived size, physical size, and informational complexity are completely different concepts.

Four Key Concepts: Subverse, Seedverse, Superverse, and Substrate

To organize these ideas, four terms prove useful (these are not standard terminology, but are defined here for consistency):

• Subverse: A created universe that depends on another reality — existing "beneath" it and subordinate to it.
• Seedverse: A universe that originates within a parent reality but subsequently becomes independent.
• Superverse: The higher-level or parent reality relative to the previous two.
• Substrate: The material or conditions upon which the created universe depends in order to exist.

If we create a simulated universe on a computer, that simulation is the subverse, our universe is its superverse, and the hardware is its substrate — whether silicon chips, optical computers, quantum processors, or future black-hole computers. The critical point: a subverse runs on top of something, making it extraordinarily fragile. Shut down the servers and time stops. Corrupt the storage and a civilization vanishes. If the superverse ends in heat death or vacuum decay, the subverse almost certainly goes with it.

A seedverse is different. It may originate from a black hole, a false vacuum bubble, or a wormhole throat — but once it "pinches off" into its own spacetime and no longer requires the parent reality to sustain it, it ceases to be a subverse and becomes a "descendant reality." A vivid analogy captures the distinction: a simulation is a lantern we must keep lit, while a seedverse is like a second candle ignited from a flame — from that point on, it burns somewhere beyond our reach.

We should also resist the temptation to conclude that anything dependent on a substrate is therefore not real. Money isn't fake just because it depends on banks and collective belief. A melody doesn't fail to exist because it depends on air vibrations. A mind isn't unreal just because it depends on neurons — or perhaps silicon. Real things can absolutely be substrate-dependent. Hints from the holographic principle go further: in some approaches to quantum gravity, a higher-dimensional spacetime is mathematically equivalent to information on a lower-dimensional boundary, suggesting that spacetime itself may be less fundamental than information and entanglement.

The holographic principle provides important theoretical context for the intuition that "spacetime depends on information." It originates from the AdS/CFT correspondence discovered by Juan Maldacena in the 1990s: a higher-dimensional spacetime containing gravity (anti-de Sitter space) is mathematically equivalent to a conformal field theory on its lower-dimensional boundary — both describe the same physical reality. This implies that gravity and particles in three-dimensional space may be a kind of "projection" of two-dimensional boundary information. One corollary: spacetime is not the most fundamental thing; quantum entanglement and information structure are the deeper substrate. This has profound implications for the question of whether substrate-dependent things are real — if our own spacetime is in some sense already an emergent phenomenon of information, then the ontological gap between a simulated universe running on computational hardware and a "real universe" may be far narrower than intuition suggests.

How to Build a Universe

Two radically different paths present themselves.

The Computational Path

The simplest artificial universes are the ones we're already building. In the future, a planet-scale computer could host vast numbers of artificial worlds, and a Matrioshka brain orbiting a star could support a staggering number of simulated environments. Even so, a simulation need not model everything — only the parts that can affect its inhabitants. Distant galaxies can remain approximations until closely observed. Quantum events can be generated statistically.

This isn't cheating. Our own universe already operates in layers: from the quantum domain to the macroscopic world to the astronomical scale, physics at smaller scales underpins emergent behavior at larger scales. When building a bridge, we don't need to solve quantum chromodynamics every time — the deeper laws haven't disappeared, they've just been compressed into "good enough" higher-level rules. A simulated universe can work the same way, increasing its resolution of detail only when events, observers, or interactions demand it.

The Cosmological Path

Here, the question isn't whether we can compute a universe, but whether spacetime can reproduce. One famous possibility involves black holes: in many speculative models, the interior of a black hole may not terminate at a singularity but may instead be the birthplace of a new, expanding spacetime.

Lee Smolin's "cosmological natural selection" hypothesis suggests that universes producing more black holes may produce more descendant universes, creating something like an evolutionary process among cosmic laws. Another possibility involves wormholes and inflationary bubbles: create a small patch of spacetime in a high-energy state and allow it to undergo inflation. From the outside it might appear microscopic; from the inside it could expand into a vast cosmos — the classic "bigger on the inside" scenario.

The real work of Edward Farhi, Alan Guth, and Jemal Guven is directly relevant here: they explicitly asked whether a universe could be created in a laboratory by forming a false vacuum bubble, and explored whether quantum tunneling could allow it to evolve into a new universe. But a persistent warning applies: if vacuum decay were triggered in our universe in the wrong way, it could be catastrophic. This is emphatically not a proposal for careless experimentation.

The fundamental difference between a physical baby universe and a computational subverse is this: a simulation requires a continuously existing substrate in the parent universe, while a baby universe may require only a single originating event. The initial seed may be extremely small, and the matter invested need not correspond to the energy perceived by inhabitants after inflation — the "accounting" of general relativity, horizons, and global spacetime structure is far subtler than a simple "box of stuff."

Inflation theory is the core physical backdrop for this path. Proposed by Alan Guth in 1980 to explain why our universe is so uniform and flat on large scales, its basic idea is that the universe underwent exponential expansion for an extremely brief period after its birth (around 10⁻³⁶ seconds), stretching a sub-atomic region to macroscopic and even cosmic scales. The driver of this process is the "false vacuum" — a quantum field state with extremely high energy density that has not yet decayed to its lowest energy state. The false vacuum has negative pressure, producing a repulsive gravitational effect that drives space itself to accelerate its expansion. Once this field decays to the "true vacuum," inflation ends, and the released energy converts into particles and radiation — the "reheating" phase. This mechanism is precisely what makes it physically conceivable, at least in principle, to artificially trigger a false vacuum bubble and allow it to inflate internally into a new universe — even though the required energy density far exceeds any foreseeable technological capability.

When a Universe Outgrows Its Creator: Ethics and the Fermi Paradox

If artificial universes are possible, the greatest questions may not be technical but moral. Suppose we create a simulated universe with conscious inhabitants who live, love, suffer, build civilizations, and study physics. What do we owe them? Are we obligated to keep their universe running? Are we permitted to modify their laws, edit their memories, reboot their history, or terminate the experiment?

Creating a world full of conscious life is nothing like building a clock. It's more like taking responsibility for a civilization that never asked to exist. If you can pause their time, erase their memories, and adjust their physical constants, then from their perspective you aren't just a programmer — you are a cosmic force with administrator privileges. That may be one of the most theologically alarming phrases imaginable.

Physical baby universes change the ethical landscape entirely. If such a universe becomes independent, we may have no ability to govern, repair, or even observe it. This is closer to parenthood than programming: parents bring a child into existence, shape its earliest circumstances, and then lose control — the child becomes a person, not a possession.

This also has implications for the Fermi paradox. We tend to imagine advanced civilizations expanding outward — colonizing planets, building Dyson swarms. But perhaps some mature civilizations turn inward: building simulated worlds, archives, afterlives, or baby universes launched beyond the observational horizon. If so, the greatest works of advanced civilizations might be completely invisible to us — from the outside, they might look like a black hole, a burst of radiation, or nothing at all. If a society only needs to launch one viable seedverse rather than colonize an entire galaxy, the motivation for endless outward expansion may diminish considerably — which actually makes this a more compelling explanation of the Fermi paradox.

In a universe facing heat death, this temptation is especially powerful. When stars burn out, black holes evaporate, and usable energy is exhausted, the ultimate escape may not be faster ships or better shelters, but cosmological reproduction.

The Fermi paradox, posed by physicist Enrico Fermi in 1950, asks: given that the universe is billions of years old and the Milky Way contains hundreds of billions of stars, advanced extraterrestrial civilizations should statistically exist in vast numbers — so why have we never detected any reliable signal or evidence of a visit? Classic explanations include: the origin of life is extraordinarily rare (the Rare Earth hypothesis), advanced civilizations tend to destroy themselves (the Great Filter theory), interstellar travel is physically prohibitive, and civilizations choose silence or concealment. The "inward turn" hypothesis raised here is a behavioral and motivational explanation: advanced civilizations may invest their resources and intellect into building interior worlds rather than broadcasting signals or expanding territorially. This resonates with concepts like "digital civilizations" and "post-scarcity societies" — when virtual or created realities are sufficiently rich, the drive to physically conquer the observable universe loses its force, causing such civilizations to statistically "disappear" from our detection horizon.

Are We Already Inside One?

If such a chain of realities exists, we must face the unsettling reverse question: are we already part of one? Not necessarily a simulation. Our universe might be a subverse running on some higher substrate, a created work sustained by a superverse beyond physics, a seedverse born from another universe that has since become fully independent, or a universe with no parents and no programmer — a place where no one waits beyond the edge of the map.

To be honest: there is currently no clear or widely accepted evidence supporting the seedverse hypothesis. Dark energy sometimes gets pulled into these discussions, because it tells us that the vacuum and the expansion of space are stranger than common sense would expect — but dark energy is absolutely not a "birth certificate from a parent universe." Standard cosmological models continue to treat it as a cosmological constant or some evolving field.

Creating an artificial universe may ultimately prove impossible, or so distant a prospect that even a Kardashev Type III civilization would regard it as myth. But the idea forces us to think carefully about what "creation" actually means. Some things we make are lanterns we must keep lit. Others are children we can only send out into the dark. Perhaps the ultimate ambition of intelligence is not merely to survive the universe, colonize it, or understand it — but to catalyze new domains where history can begin again. And perhaps, just beyond our horizon, some prior existence once did the same for us.