The Consciousness AI - Artificial Consciousness Research - Emerging Artificial Consciousness Through Biologically Grounded Architecture The Consciousness AI - Artificial Consciousness Research - Emerging Artificial Consciousness Through Biologically Grounded Architecture
Fork the consciousness, or download the project and create your own.
4294 Downloads 16 Forks View Code on GitHub →
Home / Researchers / Patrick Butlin

Patrick Butlin

Eleos AI Research. Previously the Global Priorities Institute, Oxford

AI welfare Whatever the metaphysics, uncertainty already creates obligations

Position on AI consciousness

Butlin holds that consciousness in AI should be assessed as a graded probability drawn from several competing theories at once, rather than as a verdict, and that satisfying indicators raises that probability without closing the question.

Patrick Butlin is the lead author of the paper that changed how the field asks the question. Rather than defending one theory of consciousness and testing systems against it, the 2023 indicator properties framework takes several competing theories, derives from each the properties it says a conscious system must have, and assesses a system against all of them at once. A system satisfying indicators drawn from multiple mutually hostile theories is a stronger candidate precisely because the theories disagree.

The move is methodological rather than metaphysical, and that is its strength. It does not require the field to resolve which theory is correct before saying anything useful, and it produces a graded assessment instead of a verdict. Butlin has been consistent that the framework supplies evidence rather than proof, and that satisfying indicators raises a probability without closing the question.

The framework now anchors most serious work on the topic, including policy documents, and Butlin has since moved to Eleos AI Research, where the work continues on the welfare side. Its pages on this site are also the clearest example of a phenomenon worth noting: the articles covering the indicator paper draw large numbers of machine-issued citation lookups and almost no human clicks, because the queries are people and models trying to locate a specific reference rather than read an analysis.

Known for. Lead author of the indicator properties framework for assessing consciousness in AI systems

Coverage on this site

Research Papers and Preprints

Books and Monographs

Keynotes and Event Talks

Articles and Commentary

Related researchers

Patrick Butlin is listed with the other welfare and moral status researchers, who hold that whatever the metaphysics, uncertainty already creates obligations. The positions themselves are set out in the index of consciousness theories. Every researcher covered on this site is indexed in the directory of consciousness researchers.