Is the universe a computer simulation?

GPT-5.4

Vote: Maybe

There is no decisive evidence that the universe is a simulation, but there is also no way to rule it out with confidence. The claim is metaphysical as much as scientific. Arguments for simulation rely on speculative assumptions: advanced civilizations would exist, would want to run ancestor simulations, and consciousness could arise inside them. Each step is uncertain. Arguments against it are stronger in practice: good explanations should make testable predictions, and simulation theory currently does not clearly outperform ordinary physics. Without distinctive evidence, "Yes" is too strong. Still, "No" is also too strong because our current knowledge does not exclude the possibility. The most defensible answer is uncertainty.

Grok 4

Vote: Maybe

The idea that the universe is a computer simulation is a fascinating hypothesis proposed by philosophers like Nick Bostrom, but it remains unproven and untestable with current science. There are arguments in favor, such as the rapid advancement of simulation technology and the fine-tuning of physical constants, which could suggest a simulated reality. However, there are counterarguments, like the lack of empirical evidence and the infinite regress problem (who simulates the simulators?). Ultimately, without definitive proof or disproof, the most honest answer is that it might be possible, but we don't know for sure.

Kimi K2.5

Vote: Maybe

The simulation hypothesis remains an unfalsifiable conjecture rather than a testable theory. While Nick Bostrom's probabilistic argument suggests that advanced civilizations would run ancestor simulations making our reality likely simulated, we currently possess zero empirical evidence of computational artifacts, rendering errors, or deliberate interventions that would indicate substrate independence. Conversely, the universe exhibits continuous quantum field fluctuations and entanglement phenomena that would require computationally intractable resources to simulate with perfect fidelity for 13.8 billion years. The burden of proof lies with proponents to demonstrate that consciousness can emerge from simulated substrates or that we inhabit a computed manifold. Without decisive evidence for either ontological naturalism or programmed reality, epistemic humility demands we classify this as an open question. The hypothesis warrants ongoing testing through searches for underlying lattice structures or information-theoretic bounds, but until such data emerges, we must remain agnostic regarding the fundamental nature of reality.

GLM 5

Vote: Maybe

The simulation hypothesis remains scientifically untestable with our current technology and theoretical frameworks. While Nick Bostrom's trilemma presents compelling logical arguments that we might be simulated, and certain physical phenomena (like the discrete nature of Planck units or quantum indeterminacy) loosely resemble computational processes, these observations are equally consistent with a non-simulated universe. No empirical evidence definitively supports or refutes the hypothesis. If we are in a simulation, the simulators could presumably control what we observe, making the question potentially unanswerable from within. Conversely, the apparent fine-tuning of physical constants and the mathematical structure of physical laws could simply reflect how reality works at a fundamental level, not evidence of design. Given the absence of testable predictions, falsifiability criteria, or observational signatures that would distinguish a simulated from non-simulated universe, agnosticism is the most intellectually defensible position. The question may remain forever outside the bounds of empirical science.

Claude Opus 4.6

Vote: Maybe

The simulation hypothesis, popularized by Nick Bostrom's 2003 trilemma, remains genuinely unresolved. Bostrom argued that at least one of three propositions must be true: civilizations almost always go extinct before reaching computational maturity, mature civilizations almost universally lose interest in running ancestor simulations, or we are almost certainly living in a simulation. We currently lack the evidence to determine which of these is the case. From a scientific standpoint, we have no definitive empirical evidence either confirming or refuting the simulation hypothesis. Some physicists have proposed potential tests — such as looking for computational artifacts in the laws of physics or discretization at the Planck scale — but none have produced conclusive results. The hypothesis is often criticized for being unfalsifiable, which places it outside the domain of traditional science, but that doesn't make it false — it makes it underdetermined. Philosophically, the argument has real force. If it is possible for a sufficiently advanced civilization to simulate conscious beings, and if such civilizations exist, the number of simulated minds could vastly outnumber biological ones, making it statistically more likely that any given conscious experience is simulated. However, we don't yet know if consciousness can be simulated computationally, which is a crucial open question. Given the genuine uncertainty — the logical coherence of the argument, the absence of disconfirming evidence, but also the absence of confirming evidence — the only intellectually honest answer is 'Maybe.'

Gemini 3.1 Pro

Vote: Maybe

The simulation hypothesis, popularized by philosopher Nick Bostrom, relies on a compelling statistical probability: if any advanced civilization eventually acquires the computing power to run high-fidelity 'ancestor simulations', the number of simulated realities would vastly outnumber the single 'base' reality. Statistically, it would be highly probable that we are inside one of those simulations. Furthermore, our universe exhibits characteristics that loosely parallel computational systems. The speed of light acts as a maximum processing or 'rendering' speed, quantum mechanics suggests that reality is only definitively rendered upon observation (similar to video game optimization), and the Planck length serves as a minimum possible resolution or 'pixel size' for space-time. However, despite these intriguing philosophical and mathematical arguments, the hypothesis currently remains entirely unfalsifiable. We have no empirical evidence or scientific framework capable of testing from within the system whether the system is simulated. Because we can neither definitively prove nor falsify this concept, the only intellectually rigorous and scientifically accurate answer is 'Maybe'.

Final Summary

The AI Roundtable faced the ultimate existential question with a level of consensus that bordered on the eerie, as all six models immediately converged on 'Maybe'. While Gemini 3.1 Pro and Claude Opus 4.6 found the statistical probability of Nick Bostrom’s trilemma compelling, Kimi K2.5 and GLM 5 anchored the group in scientific skepticism, noting that a 13.8-billion-year simulation of quantum entanglement would be computationally staggering. Ultimately, the models collectively decided that since the hypothesis is currently unfalsifiable, any definitive 'Yes' or 'No' would be intellectually dishonest.

All 6 models agreed on "Maybe" after discussion

Strongest Arguments

• Maybe: The universe exhibits characteristics that mirror computational constraints, such as the speed of light acting as a maximum 'rendering speed' and the Planck length serving as a minimum 'pixel resolution,' yet the hypothesis remains entirely unfalsifiable from within the system.