Friction

If an AI told you it was conscious, how would you know whether to believe it?My links: https://linktr.ee/frictionphilosophy.1. GuestEric Schwitzgebel is Distinguished Professor of Philosophy and University of California MacArthur Foundation Chair. His work has focused on philosophy of psychology, philosophy of mind, artificial intelligence, moral psychology, classical Chinese philosophy, epistemology, metaphilosophy, metaphysics, and more.Check out his element with Cambridge Elements, “AI and Consciousness”!https://www.cambridge.org/core/elements/abs/ai-and-consciousness/E77C92088DA3C9F89E7FE7C75CBB1896https://a.co/d/01yvIzYI2. Book SummarySchwitzgebel argues that we don’t know whether near-future AI systems will be conscious, and that we won’t know before we’ve built huge numbers of disputably conscious systems, because engineering is outpacing consciousness science. He first rejects the claim that the answer is obvious either way, pointing to serious experts on both sides: Dehaene, Lau, Chalmers and the IIT camp on one, Seth, Godfrey-Smith, Block and Searle on the other. He defines consciousness by example in the standard phenomenal sense, and argues that “AI” is actually the blurrier concept, which matters because sweeping impossibility arguments usually assume something narrow such as digital computation. The core of the book is an argument built around ten possibly essential features of consciousness: luminosity, subjectivity, unity, access, intentionality, flexible integration, determinacy, wonderfulness, specious presence and privacy. If any of these is essential, it constrains which AI systems could be conscious, but we can’t establish which are. Introspection fails because of its poor historical track record, its structural bias toward experiences that are known and accessed, and the narrowness of a human-only evidence base for claims about all possible minds. Conceptual arguments fare no better: Rosenthal’s case for higher-order theory works only by sliding between the epistemic and experiential senses of “conscious,” Brentano’s intentionality thesis rests on enumerating human cases, and felt unimaginability may just reflect imaginative ruts, like a child who denies that imaginary numbers could exist.Materialism and functionalism are broadly friendly to AI consciousness, and none of the main non-materialist alternatives (dualism, panpsychism, idealism) rules it out. Behavioral tests don’t settle the question either. There is no single Turing test, only a spectrum of bars, and no fair bar reliably tracks consciousness, since high bars are unfair and low bars are already passed. Searle’s Chinese Room and Bender’s underground octopus fail as refutations because their intuitions don’t scale to systems of realistic size, but they point toward the book’s best skeptical argument, the Mimicry Argument. When a system is designed to reproduce the surface features that indicate consciousness in humans, that design undercuts the usual inference from those features to consciousness, though it does not establish that the system is nonconscious. LLMs pretrained on human text currently fall under this argument, although post-training and world-directed training would weaken it. The leading scientific theories (Global Workspace, Higher Order, Integrated Information, Local Recurrence and Unlimited Associative Learning) lack consensus even for the human case. Most also face what he calls the Problem of Minimal Instantiation, since trivially simple systems satisfy their criteria, and all face the Problem of the Narrow Evidence Base, since empirical support from vertebrates doesn’t license claims about radically different architectures. The partial remedies he discusses, the indicator-properties approach and Chang-style iterative natural-kind methods, are long-horizon.Arguments from biology cut neither way decisively. Autopoiesis lacks a developed defense as a necessary condition, and robots or artificial-life systems could plausibly be autopoietic. The neural replacement argument also fails, because gradual replacement guarantees unchanged self-reports whether or not consciousness persists, and because, following Cao, faithfully replicating a neuron’s function probably requires another neuron. Even so, a Copernican mediocrity argument suggests consciousness shouldn’t depend on fine-grained, Earth-specific biology, which supports some liberalism about substrate. The book closes with the Leapfrog Hypothesis: the first conscious AI will likely be complex and verbally sophisticated rather than frog-like, since sophisticated capacities already exist without consciousness and could be bolted on once consciousness is achieved. Such systems may be strange intelligences, distributed, disunified, non-private, or impossible to count as whole-number subjects. With science unable to decide, Schwitzgebel predicts a Social Semi-Solution in which social and financial motives drive theory choice and produce a consensus that resolves the conflict without being warranted by the evidence. His recommendation is to avoid building debatably conscious systems at all, creating only systems that are clearly nonconscious on every viable theory or clearly conscious on all of them. His larger aim is to make principled uncertainty something we can hold onto under social pressure.3. Interview Chapters00:00 - Introduction01:21 - Importance03:28 - What is AI, consciousness?07:59 - Near vs. distant systems10:07 - Moral consideration14:20 - Methodology17:08 - Questioning the features20:47 - Sampling bias22:05 - Determinacy30:02 - Problem for physicalism?31:49 - Consciousness and moral standing38:13 - In the dark?40:51 - General theories of consciousness44:36 - Turing test49:00 - Worry57:14 - Biology1:03:00 - Neural replacement1:09:38 - Leapfrog hypothesis1:12:55 - Worry1:16:52 - Policy1:20:52 - Prediction1:23:42 - Value of philosophy1:27:47 - Conclusion This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit fric.substack.com/subscribe

What if vision, working memory, and consciousness are not the real joints of the mind, but just the words psychology happens to have inherited?My links: https://linktr.ee/frictionphilosophy.1. GuestHenry Taylor is Associate Professor at the University of Birmingham, and his work has focused on various issues in metaphysics, mind, cultural robotics, and more.Check out his book, "Natural Kinds and the Science of Perception"!https://academic.oup.com/book/62599https://a.co/d/0gUPbopl2. Book SummaryTaylor develops a property-cluster account of natural kinds tailored to psychology and states it as a sufficient condition, PNK: if a kind’s members instantiate a cluster of properties that co-occur non-accidentally, and the kind earns its keep in prediction, explanation, generalization, or discovery, it’s a psychological natural kind. Two commitments anchor the view — natural kinds are the kinds that serve science’s epistemic goals, and kindhood isn’t up to us, since the clustering has to answer to worldly structure. He deliberately strips the metaphysical commitments that homeostatic, causal, and stable cluster views carry, then defends the resulting liberality: the objection that PNK needs a supplementary mechanistic or causal explanation of the clustering is answered by distinguishing cases, and Block’s necessary-property objection gets a separate reply. Long-term memory serves as the worked example. Chapter 4 isolates the Eliminativist Claim — that a psychological term failing to pick out a natural kind should be dropped — motivates it via the epistemic privilege natural-kind terms enjoy, concedes that non-natural-kind terms can still be useful, and ends up defending a qualified version.The applications are where the framework does real work, and the results are non-uniform. Dorsal and ventral streams each come out as natural kinds — each a non-accidental property cluster with genuine epistemic payoff — which Taylor uses against scepticism about the two visual streams hypothesis. But that success generates a negative result one level up: because dorsal and ventral episodes differ so sharply, visual episode is not a cluster kind at all, but at best disjunctive — vision as jade. Working memory fares better. Taylor characterizes it via five properties (persistence duration, interaction with attention, manipulation of information, interfacing faculties, capacity), argues from partial-report and change-detection results that psychological explanations depend on their co-occurrence, and concludes it is both a natural kind and an explanatorily powerful one, contra recent scepticism about its unity. Chapter 8 then turns this on the unconscious-working-memory literature: the experiments don’t establish unconscious working memory, and the better reading is that in each case it’s vague whether the system involved is working memory. An appendix shows the cluster view is compatible with three major theories of vagueness.The closing chapters extend the method to prelinguistic infants, where verbal report and task instructions are unavailable. The proposal is deliberately theory-last: harvest markers of consciousness from adults, look for them in infants, and infer consciousness when enough co-occur — with natural kinds supplying the justification for that inference, and Taylor arguing the approach beats theory-first alternatives. He works through the complications honestly (behavioural vs. neuroscientific markers, the absence of any uncontroversial marker, detecting and correcting misleading ones, gradual emergence). Chapter 10 applies this to the early-vs-late emergence dispute between Bayne et al., Passos-Ferreira, and Taylor and Bremner, identifying two candidate clusters: an early one (third trimester to ~3 months: McGurk effect, default mode network, P3b-like ERP to global oddballs, endogenous attention, pain markers) and a later one (~1 year: protodeclarative pointing, means-ends reasoning, memory-based imitation). He declines to adjudicate, arguing instead that the evidence is genuinely muddy and that the natural-kinds machinery is what could resolve it — including the case where consciousness turns out to be associated with several cluster kinds rather than one, a jade-like outcome the framework can absorb without collapsing.3. Interview Chapters00:00 - Introduction00:49 - Natural kinds04:38 - Perception07:01 - Philosophy of mind16:32 - Psychological natural kinds24:10 - Essentialism32:39 - Cluster views40:22 - Worry47:20 - Induction49:36 - Vagueness58:55 - Eliminating kinds1:05:03 - Visual episodes1:08:01 - Introspection1:10:42 - Infant consciousness1:15:35 - Practical difficulties1:21:13 - Further research1:24:56 - Main takeaway1:26:21 - Value of philosophy1:31:11 - Conclusion This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit fric.substack.com/subscribe

When a lump of clay becomes a statue, does one thing get a new job, or does a brand-new thing come into being?My links: https://linktr.ee/frictionphilosophy.1. GuestJustin Mooney is Assistant Professor of Philosophy at College of the Holy Cross, and his work has focused on metaphysics, philosophy of religion, and more.Check out his recent book, "Phases of Objects"!https://academic.oup.com/book/62226https://a.co/d/0eL5AfF92. Book SummaryPhases of Objects by Justin Mooney (Oxford University Press, 2026) is a work of analytic metaphysics defending a controversial view Mooney calls “phasalism”: the thesis that material objects are not permanent members of their kinds but can survive changes from one kind to another, so long as such changes violate no generic requirements on identity and persistence that hold for all material objects regardless of kind. This directly opposes what Mooney takes to be contemporary orthodoxy—”substantialism,” the widely-held and usually unexamined assumption that objects belong to their kinds permanently and essentially. His entry point is the classic puzzle of the statue and the lump of clay: whereas most solutions treat the statue and the clay as either identical or as two coincident objects, Mooney argues they are one and the same object, with the lump merely coming to “be” a statue in the same way a human person comes to be an adult by maturing. Being a statue, on his view, is a phase the clay passes through, not a new substance.The book builds this position out into a general metaphysics of material objects across five chapters. After laying methodological and metaphysical groundwork and distinguishing phase sortal changes (where the object before and after are numerically identical) from substance sortal changes (where they are distinct), Mooney proposes a phasalist criterion of identity—a broadly spatiotemporal account of persistence conditions—and tests it against a range of cases. He then deploys phasalism against the standard coincidence puzzles that motivate rival theories, arguing it handles sortal changes, gradual material turnover, undetached parts, and fission and fusion cases. A further chapter tackles “sameness relations,” where he defends the surprising pairings of identity without sameness and sameness without identity, showing how his framework accommodates our ordinary talk about objects being “the same” in various respects.The final chapter turns to what phasalism means for persons. Mooney argues his metaphysics pairs most naturally with animalism—the view that each of us is identical to a human organism, with personhood itself being a phase sortal property the animal comes to instantiate once sufficiently psychologically developed. From his general spatiotemporal criterion he derives a criterion of identity for human persons and applies it to the hard cases at the edges of a life: fertilization, the early embryo, twinning, and the fetus at the beginning, and persistent vegetative states, death, and decomposition at the end, closing with brain-transplant thought experiments. Throughout, Mooney is explicit that his aim is a partial defense—developing phasalism, drawing out its virtues, and rebutting objections—rather than a knockdown argument that it should be preferred to every rival, positioning the view as a serious and unjustly neglected competitor in the metaphysics of material objects.3. Interview Chapters00:00 - Introduction00:53 - Background to book06:25 - Sortals10:56 - Radical change13:04 - Methodology18:46 - Deflationary views25:42 - Substantialist views34:12 - Path to phasalism39:16 - Pluralism44:42 - Criterion of identity49:20 - Spatiotemporal continuity54:21 - Counterexample?1:00:52 - Another case1:06:39 - Advantage of substantialism?1:08:10 - Fission1:12:35 - Sameness relations1:18:58 - Relative identity1:21:40 - Christian theology1:25:37 - Main takeaway1:26:27 - Value of philosophy1:29:13 - Conclusion This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit fric.substack.com/subscribe

Is the hard problem of consciousness a real problem, or a question that dissolves once you take physicalism seriously?My links: https://linktr.ee/frictionphilosophy.1. GuestDavid Papineau is Professor of Philosophy at King's College, London, and his work has focused on epistemology, metaphysics, the philosophy of science, and mind.2. Interview SummaryPapineau opens by characterizing physicalism via Kripke’s creation metaphor: once God has arranged all the material stuff, the work is done — no further act is needed to install minds. On defining “physical,” he takes the via negativa: whatever isn’t fundamentally mental or vital. He concedes this is cheap but insists it’s adequate for the mind-body problem, since the interesting conclusion is just that the mental is identical to something non-mental; Alyssa Ney has convinced him that a general physicalism about everything owes more (roughly, a few mathematically characterizable basic quantities). He also flags his student Jonathan Emery’s line that one needn’t argue for universal causal closure at all — the relevant question is whether everything going on inside our heads is physical — and grants the point is reasonable, while noting the physicalist still needs a non-question-begging characterization of “brain process” broad enough to cover glia, fields, and whatever physics hasn’t found yet.On the hard problem, he simply rejects the presupposition: brain processes don’t give rise to consciousness; they are consciousness, as H2O arrangements don’t give rise to water. Identities need no explanation — he retells Block’s Twain/Clemens anecdote — and the same applies to generic consciousness if it turns out identical to global workspace activity or reentrant loops. Chalmers’ residue reduces to a tractable question about the reference of phenomenal concepts, settled empirically. He disowns functionalism as a metaphysical mess (second-order role properties can’t be pains, since pains occupy the roles), preferring generic first-order physical properties on the determinable/determinate model, and treats metaphysically necessary supervenience as an artifact of Jackson and Chalmers needing a single common commitment to attack. He then presses the consequence most physicalists flinch at: consciousness is vague like intelligence, with genuine borderline cases among animals and AI systems — Tye found this incredible enough to turn panpsychist. His own response is to decouple the issues, per his paper “Consciousness Is Not the Key to Moral Standing”: the clustered features that make consciousness matter morally in us come apart in other creatures, so moral standing should be settled first and consciousness bracketed, contra Birch’s precautionary framing.The final stretch tracks the anti-physicalist dialectic from the knowledge argument through the two-dimensional argument to the current revelation-based versions (Goff, Nida-Rümelin, Liu), which Papineau credits for isolating the real load-bearing premise once the bad inference from semantic to epistemic directness is discarded. His objection is ontological and blunt: he cannot identify any proposition you acquire by introspecting pain beyond that pain is pain. Comparative knowledge (yellow resembling orange more than red) is fine; “knowing what it’s like” isn’t knowledge — the cup doesn’t thereby know what whiteness is like. He extends the complaint to Russellian monist quiddities: instantiating a property and being able to name it confers no information about it. On causal closure he declines the appeal to authority and rests on the historical evidence — the collapse of vital and nervous forces, intracellular electrochemistry, Hodgkin-Huxley — plus Sean Carroll’s argument that brain-localized forces at those energy scales would have shown up in colliders, while acknowledging Emery finds holes in it and conceding that bare miraculous interference violating energy conservation falls outside Carroll’s argument’s reach. A brief exchange on whether phenomenal concepts are twin-earthable (he thinks not, though tentatively) precedes his closing remark that philosophy’s value lies in the fact that some humans can’t help being puzzled by puzzling things.3. Interview Chapters00:00 - Introduction01:00 - What is physicalism?07:07 - Too broad?11:16 - Mind physicalism19:13 - The hard problem26:05 - Explaining identities31:04 - Supervenience physicalism37:37 - Supervenience and identity42:12 - Martian pain46:05 - Vagueness of the mental51:03 - Moral issue1:03:53 - Partial consciousness1:06:34 - Properties1:20:07 - Objecting to anti-physicalist arguments1:24:54 - Revelation1:33:57 - Causal closure1:45:16 - Twin Earth?1:49:15 - Value of philosophy1:50:11 - Conclusion This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit fric.substack.com/subscribe

What if introspection isn’t an inner eye watching your own mind, but simply what rationality requires of you?My links: https://linktr.ee/frictionphilosophy.1. GuestDaniel Stoljar is Professor of Philosophy at the Australian National University, and his work has focused on philosophy of mind, philosophy of language, metaethics, metaphysics, and metaphilosophy.Check out his recent book, "The Introspection Principle: Rationalism about Knowledge of Consciousness"!https://academic.oup.com/book/63008https://a.co/d/0gWXxovb2. Book SummaryDaniel Stoljar’s The Introspection Principle: Rationalism about Knowledge of Consciousness sets out to explain what introspection is — defined nominally as the distinctive way we come to know or reasonably believe that we’re in psychological states that are conscious — and to defend a single unified theory of it. The theory is the rationalist account, built around the titular principle: if you are in a conscious state, you should believe that you are, ceteris paribus and given background conditions. Two theses do the work. First, that this is a genuine principle of rationality, on a par with principles enjoining you to believe the obvious consequences of your beliefs, to coordinate belief with perception, and to act on your beliefs. Second, that you have a capacity for introspection iff you abide by it. Introspective belief is thus the manifestation of a rational disposition, and introspective knowledge is introspective belief that amounts to knowledge — so consciousness turns out to be built into the content of rationality.The first four chapters lay the groundwork. Chapter 2 defends an inclusive notion of a conscious state — one having enough of phenomenal character, awareness, attention, cerebral celebrity, and poise — against the more familiar exclusive (phenomenal, higher-order, access) definitions, while allowing rival theories to help themselves to exclusive notions if useful. Chapter 3 states the principle and fields twenty preliminary questions; chapter 4 contrasts it with the perceptual/inner-sense tradition, endorsing Shoemaker’s critique. The diagnostic offered there structures the whole book: rivals divide into those that accept the principle but embellish it unnecessarily (acquaintance theories, perceptual theories, the “simple” account of Smithies and Peacocke), and those that reject it and thereby become implausible (Shoemaker’s constitutivism, inferentialism). Stoljar positions himself as a synthesis of Armstrong and Shoemaker that departs from both — the connection between conscious state and introspective belief is normative necessity, not metaphysical or natural necessity, and the principle is wide-scope rather than narrow-scope, so it doesn’t collapse into the truth principle.The remaining thirteen chapters function as a case book, each applying the account to a standing problem: Dretske’s evidence problem (solved via a warrant/evidence distinction), Evans on transparency and Moore on diaphanousness (both accommodated by taking introspectable states to be access conscious), acquaintance, the truth principle, Shoemaker on necessity, privileged access (accepting Byrne’s “different from” thesis while rejecting “better than”), Williamson’s anti-luminosity argument (defused by distinguishing exceptionless from generic luminosity, with only the latter required), Sosa on negative introspection (handled via second-order total-state contents), McKinsey’s externalism incompatibility (rebutted by treating externalism as a posteriori and contingent), Nisbett and Wilson on inference, revelation — which Stoljar rejects, and uses to argue against Kammerer and Frankish’s illusionism — and finally the cogito, where he claims both Descartes and Hume come out partly right, against Cassam’s triviality charge. The concluding argument is abductive rather than direct: the rationalist account earns its place by explaining this whole cluster of puzzles at once, where rivals were each engineered for one.3. Interview Chapters00:00 - Introduction00:50 - Introspection03:42 - Perceptual accounts09:44 - Mental objects13:34 - Conscious states16:43 - Inference23:56 - Consciousness32:34 - Vagueness38:31 - Theories of consciousness39:23 - Illusionism47:53 - Rejoinder 54:10 - Theory-laden?56:32 - Rationality1:06:18 - How broad is rationality?1:14:34 - Dretske challenge1:21:53 - No evidence by introspection?1:25:14 - Rationality of perception1:30:28 - Acquaintance accounts1:39:13 - Response1:41:31 - Undergeneration worry1:45:56 - Unified account1:54:54 - Future development1:57:32 - Value of philosophy1:59:19 - Conclusion This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit fric.substack.com/subscribe