TLDR

  1. A first-principles research programme for studying experience, human intelligence and artificial systems without confusing fluent report, functional competence, subjecthood or moral status.
  2. Begin with why experience resists reduction to function.
  3. Then learn how to compare candidates, define the subject under test and govern consequential systems before certainty arrives.
  4. The Hard Problem and the Price of Every Exit.
  5. A Consciousness Assurance Case: Governing Machines We Cannot Classify.
Home · Writing · Consciousness

Consciousness, Minds and Machines

A first-principles research programme for studying experience, human intelligence and artificial systems without confusing fluent report, functional competence, subjecthood or moral status.

Programme figureUncertainty becomes an engineering decisionConceptual phase plot
Research path from explanatory gap to reversible governance A four-stage curve rises from the hard problem through comparative evidence and candidate-boundary tests to an assurance case. The vertical axis represents decision burden, not evidence for consciousness. highlowResearch sequence →Decision burden Explanatory gapwhat function leaves open Comparative evidencewhich observations transfer Candidate boundarywhat interventions preserve Assurance decisionwhat remains reversible research pathunresolved theoretical rangeschematic · no measured values
Finding. The metaphysical question can remain open while the engineering decision becomes sharper. Comparison, candidate-boundary tests and reversible controls increase the burden of evidence without pretending to measure consciousness directly. Conceptual plot; no empirical scale or probability is implied.
Start here

One route through the central problem

Begin with why experience resists reduction to function. Then learn how to compare candidates, define the subject under test and govern consequential systems before certainty arrives.

01 The Hard Problem and the Price of Every ExitWhy the question resists a clean verdict Philosophy, research strategy and decision framing02 A Comparative Atlas of Possible MindsHow to compare unfamiliar candidates without one ladder Comparative cognition, AI evaluation and welfare research03 Before Experience: The Minimal ConstraintsHow to define a candidate that experiments can actually test Experimental design and candidate-boundary work04 A Consciousness Assurance Case: Governing Machines We Cannot ClassifyHow to govern a system while classification remains open Engineering, governance and independent review
Research programmes

Continue by the question you need to answer

Inquiry and epistemic discipline

Five Words People Confuse: Intelligence, Consciousness, Sentience, Selfhood and AgencyIntelligence, consciousness, sentience, selfhood and agency are different claims, each with its own evidential burden and practical consequence.Three Ledgers: Function, Mechanism and ExperienceA claim about what a system does cannot become a claim about how it works or what it feels without an explicit bridge that can fail.From Observation to Explanation: Correlation, Causation and ConstitutionObservation reveals patterns, intervention can identify causes, and organised realisation is needed to explain what a phenomenon is.What Counts as Evidence for Another Mind?Testimony is a causal trace whose evidential force depends on how the report was generated and which rival generators remain live.Thought Experiments as Instruments, not OraclesA thought experiment earns force by exposing dependencies among premises, bridges and countermodels, not by making one intuition feel decisive.Levels, Boundaries and the Grain of ExplanationThe useful explanatory grain is the smallest boundary and timescale that preserves the intervention relevant to the decision.First-Person, Interpersonal and Third-Person EvidenceReports, elicitation, behaviour, physiology and mechanism constrain one another only when each transformation and bridge remains explicit.How to Read a Frontier ClaimA frontier claim is a chain from observation through interpretation to action, and each link must be read and tested separately.

Mathematical language and representation

Number, Measurement and RepresentationA property becomes actionable through a representation that determines which comparisons, transformations and decisions are legitimate.Functions and Compositional Behaviour: Why the Model is the Path, not the PileParameters become behaviour only through the ordered, routed and repeated composition of transformations that forms an executable path.Sets, Logic and ProofA proof strengthens a claim by eliminating every admissible countermodel left open by its domain, relations, premises and modal frame.Counting and Combinatorial ExplosionSmall independent choices multiply partial worlds, so tractable systems survive by proving most candidates impossible, equivalent or irrelevant.

Experience, theories and candidate subjects

Before Experience: The Minimal ConstraintsCandidate definition and experimental minimumsBinding and the Candidate-Subject BoundaryUnity, integration and multi-agent subject boundariesThe Hard Problem and the Price of Every ExitThe explanatory gap and the price of each exitResearch Without Theory VictoryProgress when rival theories remain liveIntegrated Information and the Causal Structure of Artificial SystemsCausal structure and integrated-information claimsBuilding an Attention Schema Without Mistaking Report for ExperienceAttention schemas, report and interventionDoes the Substrate Matter, or the Causal Organisation?Substrate dependence and causal organisationEmergence and the Ruler that Creates ItScale, emergence and measurement choiceA Comparative Atlas of Possible MindsA multidimensional map of possible mindsRights, Forks and Interrupted MindsContinuity, branching, deletion and provisional rights

Machine Consciousness Research

A Consciousness Assurance Case: Governing Machines We Cannot ClassifyOperational assurance under moral uncertainty

Additional consciousness studies

Interoception and Grounding Why Embodiment Matters for Models of MindA precise look at interoception, from insula anatomy and Damasio's contested somatic marker hypothesis to predictive processing and contemplative body-taxonomies, and what any of it actually implies for disembodied AI agents.Machine Attention and Human AttentionTransformers made 'attention' a household word in AI. But the word carries two very different meanings. Where the analogy to biological attention illuminates, and where it quietly breaks.Machine Attention and Human Attention a Careful DisanalogyTransformer attention and human attention share a word and almost nothing else; this piece works through the mechanism precisely, then shows exactly where the analogy breaks and where, more modestly, it does not.What Predictive Processing and Kashmir Shaivism Agree AboutPredictive processing and the Pratyabhijna school of Kashmir Shaivism both treat the ordinary given world as constructed rather than found, but they are built to answer different questions, and the popular claim that one proves the other is a specific, nameable error.