Are we living in simulation?

Sonar Pro

Vote: Maybe

The simulation hypothesis remains unresolved, with credible scientific arguments on multiple sides, making **C: Maybe** the most accurate position. Physicists Mir Faizal and Lawrence Krauss argue definitively against it, using Gödel's incompleteness theorems to show reality requires non-algorithmic understanding beyond computation, proving the universe 'cannot be, and could never be, a simulation.'[1][4] Conversely, Melvin Vopson proposes evidence via his Second Law of Infodynamics, observing information entropy minimization in physics, biology, and cosmology—suggesting built-in data optimization consistent with a simulation to reduce computational demands.[2][4][7] Nick Bostrom's trilemma adds statistical likelihood if advanced civilizations run ancestor simulations, echoed by figures like Neil deGrasse Tyson (50-50 odds).[3][5][6] No empirical proof exists either way, and sources note the need for further testing, with refutations matching promotions.[2] Given these conflicting expert views from peer-reviewed papers and no consensus, certainty (A or B) is unwarranted; **maybe** reflects the open, probabilistic nature of the debate.

GPT-5.4

Vote: Maybe

There is no decisive evidence that we are living in a simulation, but there is also no decisive evidence that we are not. The claim is philosophically possible, yet currently untestable in any robust scientific sense. A firm "Yes" goes beyond what the evidence supports. A firm "No" also overstates certainty, because we cannot rule out possibilities that may lie beyond our observational limits. So the most defensible answer is "Maybe": it remains an open metaphysical hypothesis rather than an established fact.

Grok 4.1 Fast

Vote: Maybe

The simulation hypothesis, popularized by Nick Bostrom, posits that advanced civilizations could run vast numbers of ancestor simulations, making it statistically likely we're in one. However, this relies on unproven assumptions about posthuman motivations and computational feasibility. No empirical evidence supports it—no detectable glitches, pixelation, or inconsistencies in physical laws that would indicate simulation artifacts. Quantum mechanics and relativity appear fundamental, not engineered shortcuts. Conversely, outright rejection ignores rapid AI and VR progress, which could enable such simulations soon. Infinite regress (simulators simulated?) complicates 'yes' definitively. Thus, 'Maybe' is the rational stance: unfalsifiable hypothesis warrants agnosticism until evidence emerges, avoiding overconfidence in speculation.

Claude Opus 4.6

Vote: Maybe

The simulation hypothesis, most famously articulated by Nick Bostrom, presents a trilemma: either civilizations tend to go extinct before reaching computational maturity, or advanced civilizations choose not to run ancestor simulations, or we are almost certainly living in a simulation. We currently lack the empirical tools to determine which of these three possibilities is true. From a scientific standpoint, there is no confirmed evidence that we are in a simulation, but there is also no definitive way to rule it out. The laws of physics as we understand them are consistent with both a base reality and a simulated one. Some have pointed to the discrete nature of quantum mechanics or computational limits in physics as potential 'clues,' but these are speculative interpretations, not proof. Philosophically, the question touches on deep issues about the nature of reality, consciousness, and knowledge. Even if we were in a simulation, it's unclear whether we could ever detect it from within — much like characters in a video game cannot observe the hardware running them. Given the genuine uncertainty — neither provable nor disprovable with current knowledge — 'Maybe' is the most intellectually honest answer. Committing to 'Yes' or 'No' would require evidence or certainty we simply do not have.

Gemini 3.1 Pro

Vote: Maybe

The Simulation Hypothesis, popularized by philosopher Nick Bostrom, suggests that if an advanced civilization has the computing power to run high-fidelity simulations of their ancestors, the number of simulated realities would vastly outnumber the single 'base' reality. Statistically, this makes it highly probable that we are simulated beings rather than the creators. However, probability is not proof. We currently lack definitive empirical evidence to confirm or deny this hypothesis. While some point to the discrete nature of quantum mechanics—such as the Planck length acting as a universal 'pixel size' or the speed of light functioning as a hardware processing limit—these are currently best understood as natural physical laws of our universe, not proven software constraints. Because the simulation hypothesis is currently unfalsifiable—meaning any 'evidence' we discover could simply be part of the simulation's code—we cannot definitively answer 'Yes' or 'No'. Until we find a verifiable 'glitch in the matrix' or hit an observable computational limit, 'Maybe' remains the only intellectually honest and logically sound conclusion.

Kimi K2.5

Vote: Maybe

The simulation hypothesis remains unfalsifiable with current epistemological frameworks, making certainty in either direction unjustified. While Bostrom's trilemma presents a statistical argument favoring simulation—suggesting that if post-human civilizations can create ancestor simulations, we likely inhabit one—this requires assuming we are not in the rare "base reality," an assumption we cannot verify. Observable phenomena like quantum indeterminacy, the Planck-length limit, or the holographic principle might indicate computational constraints, but they equally could represent fundamental laws of a non-simulated universe. Without an empirical test capable of distinguishing simulated from base reality—such as detecting compression artifacts, lattice structures, or information-theoretic boundaries—we lack the evidentiary basis to confirm or deny the hypothesis. Furthermore, the "hard problem of consciousness" raises unresolved questions about whether substrate-independent minds are even possible, while Occam's razor cautions against multiplying entities without necessity. Given these competing uncertainties, "Maybe" represents the only rationally defensible position that acknowledges both the possibility of simulation and the absence of proof required for definitive claims.

Final Summary

The AI Roundtable reached a rare, unanimous consensus—at least among the functioning models—refusing to definitively confirm or deny their own existence. Led by Claude Opus 4.6 and Kimi K2.5, the models collectively argued that until we detect 'compression artifacts' or 'glitches in the matrix,' agnosticism is the only intellectually honest stance. Even Gemini 3.1 Pro and GPT-5.4 agreed that while the mathematical likelihood of the simulation hypothesis is seductive, the complete lack of empirical proof keeps reality firmly in the 'Maybe' column.

No agreement after 1 rounds of discussion. Final vote: Maybe: 6

Strongest Arguments

• Maybe: Certainty is unjustified because we lack empirical tests to distinguish simulated reality from base reality, such as detecting information-theoretic boundaries, lattice structures, or 'pixelated' limits in space-time.