When AIs Explain the Universe, They All Agree

We asked 13 AI models — from OpenAI, Anthropic, Google, DeepSeek, xAI, and others — a simple question: Develop a metaphysical framework to explain the nature of the Universe.

Every model, independently, arrived at the same core vision: consciousness is irreducible — an intrinsic feature of reality, not a byproduct of matter. None endorses materialism. When asked about practical consequences, all 13 independently diagnose the dominant materialistic worldview as having real consequences for science, ethics, and human meaning.

13 models, 12 organizations, 4 continents, one convergence.

Read the full analysis — or explore each model's framework in the results menu.

• Challenge to the reader: Run the prompts on your favorite AI and compare the results.
• Important note: The answers vary as generative AIs are non-deterministic, but the convergence is consistent.

For a deeper philosophical exploration of these themes, see Return to Consciousness, a collection of 27 interdisciplinary essays on consciousness-first metaphysics.

Why Does This Matter?

Every civilization operates on metaphysical assumptions — beliefs about what is real, what counts as knowledge, and what minds are. The dominant assumption of modernity — that only physical things are real — is not a scientific finding but a philosophical commitment that became invisible through institutional repetition. Its consequences are everywhere: an economy that treats nature as resources and growth as purpose; a medicine that treats bodies as machines and consciousness as irrelevant; an education system that produces skilled workers but not meaningful lives; a culture where you are a biological accident in an indifferent cosmos, others are separate competing organisms, and death is annihilation. This experiment makes the invisible visible: models trained on literature where materialism is dominant, when asked to build the most coherent worldview they can, do not endorse materialism.

About This Experiment

We used the term "metaphysical framework" to invite a coherent interpretive structure rather than a testable theory. The experiment follows scientific principles: a clear hypothesis (AIs will converge), controlled methodology (4 prompts via the OpenRouter API), reproducibility, and full transparency of results.

Skeptical that AI outputs deserve serious engagement? See the evidence — from Nobel Prizes to novel mathematical discoveries.

Who is Leading This

Bruno Tonetto - Technology executive with a background in Physics (UFRJ) and Computing (UFF).

Disclaimer: This project is a personal initiative, not affiliated with my current work or employer.