Beyond Consciousness: Why AGI Never Feels
Zenodo (CERN European Organization for Nuclear Research) July 5, 2026 DOI: 10.5281/zenodo.19120477 via OpenAlex
Summary
AI-generated from the abstractUnder current and foreseeable computational paradigms, no known pathway leads from functional information processing to subjective experience. Whether a future non-biological system could possess consciousness remains an open metaphysical question, but the burden of proof lies with those who claim it is possible. The essay distinguishes a weak version (no known pathway, radical changes required) from a strong version (consciousness impossible in any non-biological system ever), defending only the weak version. Drawing on the hard problem of consciousness, the ontogeny argument, and the paradox of unconscious power, it argues that intelligence without a witness is likely and more dangerous than any evil AGI scenario.
Study at a glance
| Characteristics | Theoretical or philosophical paper Peer reviewed |
|---|---|
| Keywords | Consciousness Process computing Artificial general intelligence Cognition Class philosophy |
| Key finding | No known pathway from functional information processing to subjective experience exists under current computational paradigms, and artificial consciousness would require being born, not compiled or uploaded. |
Abstract
This essay argues that under current and foreseeable computational paradigms, there is noknown pathway from functional information processing to subjective experience. Whethera future, non-biological system could possess consciousness remains an open metaphysicalquestion, but the burden of proof lies with those who claim it is possible. The essaydistinguishes a weak and a strong version of the thesis: the strong version (consciousnessis impossible in any non-biological system, ever) is not defended; the weak version (noknown pathway, and radical changes in approach would be required) is. Drawing on thehard problem of consciousness (Chalmers), the ontogeny argument, and the paradox ofunconscious power, the essay shows that intelligence without a witness is not only possiblebut likely — and more dangerous than any “evil AGI” scenario. The third, expandedversion adds a comparative analysis of industry leaders with primary sources, engagementwith enactivism (Thompson, Varela) and predictive processing (Seth, Friston), formalizednecessary conditions for consciousness, and a strengthened irreversibility argument. Theessay concludes that if artificial consciousness is ever to exist, it will not be compiled oruploaded — it will have to be born.