- Wolpert proves mathematically that a universe can simulate itself exactly.
- The UBC team uses Gödel's incompleteness to argue full simulation is impossible.
- Both frameworks are internally consistent but rest on different assumptions.
Twenty-three centuries ago, the Chinese philosopher Zhuangzi woke from a dream in which he had been a butterfly. The experience unsettled him.
He could not determine whether he was a man who had dreamed of being a butterfly, or a butterfly now dreaming of being a man.
David Wolpert, a physicist at the Santa Fe Institute, opens his new paper with that same paradox. But Wolpert, known for his "No Free Lunch" theorems in machine learning, wasn't content to leave the question in the realm of philosophy. He wanted mathematics.
The simulation hypothesis has floated through physics and philosophy for decades, typically as a thought experiment rather than a formal claim. Could our universe be a program running on some advanced civilization's computer? The question sounds like science fiction.
Some physicists, notably Sabine Hossenfelder, have dismissed the hypothesis as untestable and therefore outside the domain of science entirely. Wolpert, with characteristic bluntness, observes that "this entire debate lacked basic mathematical scaffolding."
He set out to build it.
Key figure
2
Identical versions of you in a perfect self-simulation – with no experiment to tell them apart
A Program That Prints Itself
Wolpert's framework, published in the Journal of Physics: Complexity, treats universes as computational systems. He defines precisely what it would mean for one universe to simulate another, using something called the Physical Church-Turing thesis: the assumption that any measurable property of a universe can, in principle, be computed.
The stranger result emerges from a theorem most physicists never encounter. In 1938, the logician Stephen Kleene proved something peculiar about self-referential programs. A program can be constructed that outputs its own complete description. It can, in effect, print itself. This result, known as Kleene's recursion theorem, remains a cornerstone of computability theory.
What is Kleene's recursion theorem?
It is a result from the 1930s showing that a computer program can be written to output its own complete source code. This is not a trick – it follows from deep properties of how computation works. Wolpert uses it to argue that if programs can describe themselves, a universe treated as a computer might be able to generate a complete description of itself too.
This sounds like a parlor trick until you extend it, as Wolpert does, to entire universes. If a program can print itself, why couldn't a universe, treated as a computational system, compute itself?
Within his formal model, if a universe satisfies certain conditions, it can simulate itself. Not approximately. Exactly.
The theoretical construction works like this: the universe contains a computer, that computer runs a program, and the program outputs the complete future state of the universe, including the computer, including the program, including the output. Whether any physical universe could actually implement such a scheme remains an open question. Wolpert is mapping logical possibilities, not engineering blueprints.
It turns out the structure beneath the idea is richer than anyone realized.
David Wolpert, Santa Fe Institute
But those logical possibilities include you.
Which Version of You Is Reading This?
Here the identity question sharpens into something uncomfortable.
In such a self-simulation, there would be two identical versions of you. One exists in the "base" universe, running the simulation. One exists inside the simulation, experiencing identical physics and identical thoughts.
Both versions remember the same past. Both are reading this sentence at the same moment. Both believe themselves to be the original.
Which one is "really" you?
Wolpert's mathematics offers no answer. The framework treats both versions as equally real.
The question of which is the "true" you may, his analysis suggests, be meaningless. There is no experiment either version could perform to determine their status.
You could, in principle, run a simulation that includes yourself running that very simulation, and so on. It could be like staring into a mirror facing a mirror, when the reflections go on forever.
Intriguingly, Wolpert also proves that certain questions about such simulations are formally undecidable within his framework. Using Rice's theorem, he demonstrates that no algorithm can determine whether an arbitrary universe is capable of simulating itself. The question is not merely hard. It cannot be resolved by any computational procedure.
The Competing Framework Says Simulation Is Impossible
Weeks before Wolpert's paper appeared, a team at the University of British Columbia published a mathematical argument reaching the opposite conclusion, though from very different starting assumptions.
Dr. Mir Faizal and colleagues, including physicist Lawrence Krauss, invoked different mathematical machinery. Where Wolpert used Kleene, they used Gödel.
Kurt Gödel's incompleteness theorems, published in 1931, revealed a fundamental limitation of formal systems. Within any sufficiently powerful logical framework, there exist true statements that cannot be proven using the rules of that framework.
The classic example is almost playful: "This statement is unprovable." If it were provable, it would be false, creating a contradiction. If it cannot be proven, it is true, but the system cannot reach it.
The UBC team applies this insight to physics itself. They argue that reality requires what they call "non-algorithmic understanding," a form of comprehension that no sequence of computational steps can replicate. Some truths about the universe, on their view, are Gödelian. They are real but unreachable by any algorithm.
The fundamental laws of physics cannot be contained within space and time, because they generate them.
Lawrence Krauss, co-author
Since any simulation must follow programmed rules, Faizal argues, it cannot capture these non-algorithmic aspects of reality. "A fully consistent and complete description of reality cannot be achieved through computation alone," he states. The universe, on this account, cannot be fully captured by any computational process.
The philosophical leap from Gödel's theorems to claims about physical reality is contested terrain. Not all physicists accept that mathematical incompleteness translates directly into limits on what can be simulated. But Faizal's team presents it as a decisive argument.
Two Frameworks, One Ancient Question
These papers do not directly contradict each other in a strict mathematical sense. They begin from different assumptions about the relationship between physics and computation, and they end up in different places.
Wolpert assumes the Physical Church-Turing thesis holds. Every physical process, on this view, can in principle be reproduced by a computer program. He explicitly notes his analysis does not address quantum effects or the possibility of "hypercomputation" beyond Turing machines.
Faizal's team argues from quantum gravity, where spacetime itself emerges from deeper structures of information. At this foundational level, they contend, something non-algorithmic is required. The two frameworks do not directly engage each other.
Reading them side by side, as we do here, reveals something interesting. The question of whether we live in a simulation may not have a single mathematical answer. It may depend entirely on which assumptions about computation and physics you accept.
Both frameworks appear internally consistent. Neither refutes the other. They are, in a sense, answers to slightly different questions.
Zhuangzi would perhaps recognize the situation. The butterfly question remains open not because we lack rigor, but because mathematics itself forks depending on which assumptions you accept. The dreamer and the dream remain entangled.
"You think you're asking a simple question," Wolpert observes. "But once you formalize it, an entire landscape of new questions opens up."
The question of which version of you is reading this sentence may be one of them.
Sources
- Primary Research: Wolpert, D.H. (2025). What computer science has to say about the simulation hypothesis. Journal of Physics: Complexity, 6(4). DOI: 10.1088/2632-072X/ae1e50
- Additional Context:
Fact Check: Claim-by-Claim Verification Verified
The Science Reader feature article accurately represents the key claims, authors, theorems, and conclusions from both primary papers and their institutional press releases without factual errors or misattributions.
Commentary
- Both papers published in 2025; Wolpert's (Dec 1) after Faizal's (Oct), but no direct engagement between them.
- Philosophical implications like identity in self-simulations are faithfully recapped from sources without exaggeration.
Sources used for verification
Academic/Peer-reviewed:
- What computer science has to say about the simulation hypothesis - iopscience.iop.org
- Consequences of Undecidability in Physics on the Theory of Everything - jhap.du.ac.ir
Other reliable sources:
- New mathematical framework reshapes debate over simulation hypothesis - santafe.edu
- UBCO study debunks the idea that the universe is a computer simulation - news.ok.ubc.ca
- Kleene's Recursion Theorem - plato.stanford.edu
- Gödel's Incompleteness Theorems - plato.stanford.edu
Fact-checked by Perplexity Sonar Pro on 2025-12-22
