NOTE-188: The Weirdness of the World

Contribution

The book assembles eleven arguments, mostly published before as separate articles, into one thesis. It uses “philosophy that opens” (p. 8) rather than philosophy that closes. On foundational questions of mind and cosmos, every well-developed option is bizarre and none is compelling, so the right attitude is credence spread across weird possibilities. The new synthesis is chapter 3. It argues from the history of philosophy and from the incoherence of common sense that no non-bizarre metaphysics of mind exists (Universal Bizarreness). It argues from expert disagreement, the lack of any method and cosmological uncertainty that no option deserves credence much above 50% (Universal Dubiety). The other chapters work as case studies that feed the two theses. The only original data are the three self-experiments against solipsism in chapter 6.

Key insight

Common sense is not a tiebreaker in the metaphysics of mind, because common sense is incoherent there. It accepts that one of materialism, dualism, idealism or a compromise/rejection view must be true, and it rejects each of them once its implications are drawn out (p. 86). So bizarreness cannot rule a theory out, and all that remains is weighing bizarre and dubious options against each other. That licenses taking seriously group consciousness, simulation, transcendental idealism and infinite causal reach, each with a modest credence.

Assumptions

Premises and authorities, since this is a book of informal argument.

Working materialism, held conditionally. Most chapters assume “Standard-Issue Naturalism”, which is scientific materialism plus Big Bang cosmology (p. 316). The author’s own credences are about 55% materialism, 35% compromise/rejection, 5% dualism and 5% idealism (n. 134).
Definitions (ch. 3 §2). A position is bizarre if “most people without specialized training on the issue confidently, but perhaps implicitly, believe it to be false” (p. 65, after Cicero’s sensus communis). It is wild if it is bizarre and “we are not epistemically compelled to believe it”. A topic is a wilderness if something wild must be among its core truths (p. 66).
Rational credence, not knowledge. Chapter 4 never uses “know” (p. 117). Its currency is degree of belief, with numbers used “for convenience only” (p. 116).
An innocent concept of consciousness (ch. 8). It is defined by positive and negative examples. It commits only to (a) there being a single obvious category that nonspecialists latch onto and (b) realism about that category (p. 232).
Cosmological premises (ch. 7). (1) The universe is infinite. (2) It continues after heat death. (3) Galaxy-sized fluctuations recur with a finite, non-decreasing probability. (4) Actions launch infinite “ripples” of perturbed particles. Also, quantum indeterminacy is intrinsic and not observer-dependent (pp. 196–197). The chapter calls these “the most straightforward interpretation of current physical theory” (p. 206), which is its own assessment.
Two ethical assumptions (ch. 11 §2). AI personhood is possible in principle (argued in Schwitzgebel & Garza 2015). Consciousness is crucial: an AI that entirely lacks consciousness “is far short of being a person” (Claim A), and one with a fully humanlike conscious range “is a person” (Claim B) (p. 293).
Ground rules against solipsism (ch. 6 §2). The author allows himself introspection, memory from the start of the experiments only, concepts shorn of any external grounding, and the general tools of scientific reasoning (p. 170).

Key results

What each chapter argues. All of it is informal argument except chapter 6’s small self-experiments.

Ch. 1, “In Praise of Weirdness.” On the fundamental structure of the cosmos and how consciousness fits into it, the answers are beyond our current capacity to know, but “Whatever the truth is, it’s weird” (p. 4). Theories on these topics are “strange, tentative, and modish”. The chapter also defends philosophy “that opens”, which adds possibilities, against philosophy that closes (pp. 7–8).
Ch. 2, “If Materialism Is True, the United States Is Probably Conscious” (= LIT-159; published ch. 3). The chapter builds a chain. Sirian supersquids and Antarean antheads show that spatial scatter and composition from sensing parts do not block consciousness. Squidbits and “antrabbits” bring the intelligence down to rabbit level. Viewed “telescopically”, the US has every mark materialists cite: information exchange (a brain has about 10^14 synapses, against about 10^14 person-to-person “visual connections” if 1/300 of visual stimulation counts, p. 27), goal-directed responsiveness, self-monitoring, self-representation and historical embedding. The conclusion is conditional and “gappy” (p. 14). The exits are eliminativism, extreme sparseness (language required) and neurochauvinism, and each is “no escape from the quicksand of weirdness” (p. 33). The appendix answers six objections: anti-nesting (Putnam; IIT’s Exclusion Postulate), Clark’s synchrony, Dennett’s infelicity, Chalmers’s within- versus between-subsystem principle, Dretske and Kammerer on derived representation, and methodology. Each escape is said to have counter-consequences as odd as the conclusion. The main one is that under Exclusion, one extra mail-in ballot could extinguish every voter’s consciousness (pp. 45–46).
Ch. 3, “Universal Bizarreness and Universal Dubiety.” Bizarreness is argued abductively. Every well-developed metaphysics in history defies common sense, and the best explanation is that common sense about metaphysics is incoherent (pp. 61–64). Materialism faces a trilemma over pain (p. 76). Neural chauvinism means no Martian pain. Simple functionalism means beer-can pain and no mad pain. Hybrid, normality-based views lose locality: “anaesthesia by genocide”, or faster-than-light self-knowledge if what counts as normal is current and distant (pp. 74–76). Dualism faces a dilemma over causal powers and a quadrilemma over the scope of mentality: only humans, everything, a sharp line, or graded ensoulment, each bizarre (pp. 78–81). Reid and Descartes are examined as counterexamples and fail (pp. 81–83). Idealism is bizarre on its face. Compromise/rejection views (Kant, Russell, Chalmers; “no fact of the matter”) are bizarre too. Dubiety rests on three arguments: expert disagreement (§8), no good method, since “there is no conscious-ometer” (§9, p. 92), and cosmological dubiety (§10). Under dubiety no coarse position “deserves credence much over 50%”, and no moderately specific variant comes close to 50% (p. 86).
Ch. 4, “1% Skepticism.” Credence in the disjunction of radically skeptical scenarios should be about 0.1–1%, “plus or minus an order of magnitude” (p. 116). The grounds are dreaming (about 99.9% awake; for example, 20% × 20% gives a 4% credence that mundane experience is common in dreams, p. 105), a small or unstable simulation (about 99.9% not, p. 110), cosmological scenarios such as Boltzmann brains and non-benevolent creators (about 0.1%, p. 114), and a wildcard. Self-undermining evidence does not defeat the view, by Montaigne’s rhubarb (pp. 117–119). The view has practical effects. The dream-flying leap has an expected value of +0.05 against −0.003 (p. 121). Borges scores 2.995 against 2.980 for weeding, a small tilt toward short-term goods (p. 124). Remote Pascalian outcomes are set aside by symmetry, diminishing returns and loss aversion (pp. 125–127). Chapter 4 also converts the author from atheism to agnosticism (§11) and gives “I think there’s about a 99.8% chance you exist” (§12).
Ch. 5, “Kant Meets Cyberpunk.” This is a how-possibly argument for transcendental idealism about space. It holds that things in themselves are non-spatial and unknowable. An immaterial Cartesian soul (“Angel”) can implement a Turing machine: states are moods, symbols are imagined pitches, and the tape is memory associations between integers and pitches (pp. 148–152, Fig. 1). If implementation is substrate-flexible, a simulated spatial world could run on a non-spatial, non-material base, so transcendental idealism is live. Four reasons are given for a non-trivial credence (pp. 158–159). The idealism is also skeptical. Toy or random simulations would be small or unstable, which reinforces ch. 4 (pp. 159–161).
Ch. 6, “Experimental Evidence for the Existence of an External World” (with Alan Tonnies Moore). Radical solipsism predicts no evidence of anything that out-computes, out-remembers or out-plays the experiencer. Three experiments, run in one day in 2012 (n. 240), test this. (1) Primality of 20 four-digit numbers. The author’s guesses were right 11/20, and “Excel” was right 19/20 (p < .02; later p = .008). The Roman-numeral version was 13/20 against 19/20 (p ≈ .05; later .044). Notes 234 and 236 say the lone “Excel errors” were probably the author’s own. (2) Memory. The author recalled 7/20 three-digit items (recognised 14) and 8/20 three-letter items (recognised 18). The re-presented lists matched the claimed generators, 1/2012 and alternate letters of “I Have a Dream”. Recall against the confirmed lists gives both p’s < .001, and the chance of a post-hoc fit is about 1 in 10^6 (n. 241). (3) Speed chess. The author lost 17 of 20 games, with one stalemate, so 17/19, p < .001 (p. 186). Solipsism survives only with extra skepticism or increasingly ad hoc laws. The conclusion is that the external world exists, but almost nothing is established about its nature (pp. 190–191).
Ch. 7, “Almost Everything You Do Causes Almost Everything … or Infinite Puppetry” (with Jacob Barandes). Given the four cosmological premises, any hand-raise launches at least one infinite ripple. The “n^m strategy” and a convergent toy model in n. 265 put the chance that all ripples stop at about 1 in 10^100. So every finitely probable, non-unique, less-than-galaxy-sized event will eventually be caused by it. On consequentialism, every act’s utility is ”∞ + −∞” (p. 208). The same premises support signalling to far-future counterparts and “puppetry”. If the universe is also infinite into the past, as the Copernican principle suggests, we are probably puppets ourselves (pp. 213–215).
Ch. 8, “An Innocent and Wonderful Definition of Consciousness” (published: “Consciousness, Innocent and Wonderful”). “Consciousness” cannot be defined analytically, functionally or by synonym. It can be defined by example. The positive examples are sensory and somatic experience, imagery, emotion, conscious thinking and desiring, dreams, and other people’s experience. The negative examples are fingernail growth, lipid absorption, the standing disposition to answer “24”, masked 30 ms stimuli and dreamless sleep (pp. 224–227). Contentious cases, such as experience outside attention or self-awareness, are left open, and that openness is what makes the concept “Wonderful” (p. 230). Illusionist complaints, which Frankish is reported to accept, target inflated definitions, not this one (n. 286).
Ch. 9, “The Loose Friendship of Visual Experience and Reality.” The chapter defends “Multiple Veridicalities”: qualitatively different experiences can present the same distal properties accurately (pp. 240–241). The evidence is that skilled drivers are not deceived by convex mirrors (De Vos 2000; Wierwille et al. 2008), the imagined culture of Convexia, progressive lenses, and inverting lenses. Stratton and Harris are read as reporting that things keep looking “toovy” but stop looking wrong (pp. 244–247). The conclusion is that “Wherever multiple veridicalities are possible, the mind-independent world recedes” (p. 248). That is a partial, empirically informed step toward ch. 5.
Ch. 10, “Is There Something It’s Like to Be a Garden Snail?” The chapter surveys Cornu aspersum: about 60,000 central neurons and about a quarter million peripheral neurons, mostly in the tentacles (p. 254), with learning, multimodal integration and complex mating. It then argues that yes, no and gong (no determinate answer) each have antecedent plausibility. Yes rests on a slippery slope adapted from Chalmers. No rests on the models of dreamless sleep, toy robots and the enteric nervous system, and on luminosity and subjectivity (higher-order views). Gong rests on a broken concept or on vagueness, though “in-between consciousness” is not yet a developed theory (pp. 259–268). Sparseness has five dimensions: entity, state, modality, width and property type (pp. 268–271). Two things turn antecedent plausibility into posterior plausibility. One is the Common Ground Problem: theories are built on prior assumptions about prevalence. IIT is “very nearly panpsychist”, and the unlimited-associative-learning criterion is “more stipulated than argued for” (pp. 273–277). The other is the species-specificity of introspective and verbal evidence (pp. 279–281). Conclusion: “completely baffled by the case of the garden snail” (p. 282).
Ch. 11, “The Moral Status of Future Artificial Intelligence: Doubts and a Dilemma.” It defends four theses (p. 285). (1) Debatable personhood is a likely outcome of AI development. (2) It may arrive soon. The chapter uses a GPT-3-based, embodied “GPT-6 mall cop” as the near case, though it treats that case as a consensus non-person. (3) It creates the Full Rights Dilemma. Deny rights and risk mass wrongs; with a 15% credence, deleting a system is like a die roll that kills someone on a 1. Grant them and risk sacrificing humans, for example saving six robots over five humans, and undermining AI safety through “boxing” tests and the right to switch systems off (pp. 302–305). Credence-weighted rights “only partly mitigates” either horn (p. 306). The Design Policy of the Excluded Middle (with Garza) is to avoid creating AIs whose personhood is unclear, and it is conceded to be very restrictive (pp. 306–307). (4) Subhuman, superhuman and divergent AI make it worse: dog-level legal status, a “planet-sized orgasmatron”, “fission-fusion monsters” and the problem of votes (pp. 307–311).
Ch. 12, “Weirdness and Wonder.” The chapter argues for “disjunctive metaphysics”. Opening possibilities counts as progress, and exploring unanswerable questions “ranks among the most intrinsically valuable human activities” (p. 316). It also summarises how each chapter probes the limits of Standard-Issue Naturalism (pp. 316–318).

Claims

id	claim	strength	support
C1	Every well-developed metaphysics of mind is bizarre (Universal Bizarreness)	moderate	abductive argument from the history of philosophy plus four case families (materialist trilemma, dualist dilemma and quadrilemma, idealism, compromise views); an open challenge “find me one” (p. 63), not a survey
C2	Common sense is incoherent in the metaphysics of mind	weak–moderate	offered as the best explanation of C1 (pp. 61–62, 85–86); no independent test
C3	No coarse metaphysical position deserves credence much over 50% (Universal Dubiety)	weak–moderate	arguments from disagreement, lack of method and cosmological dubiety (§§8–10); the number is a judgment
C4	If materialism is true, the US is probably conscious	weak–moderate	chain of thought experiments and a Fermi estimate; conditional and “gappy” (p. 14); see LIT-159
C5	Anti-nesting principles (Putnam, IIT Exclusion) have consequences as odd as group consciousness	moderate	reductio cases (Block’s tiny people, the one-ballot threshold, the sleeping-Max flicker, n. 56); no formal Φ calculation
C6	A rational credence in radical skepticism is about 0.1–1%	weak	informal credence-multiplying per scenario (pp. 105, 110, 114); the author disclaims precision
C7	1% skepticism rationally tilts choices toward short-term goods and supports agnosticism	weak	toy decision matrices (pp. 121, 124, 133); illustrative numbers
C8	Transcendental idealism about space is a live possibility	moderate as a how-possibly claim; weak as a credence claim	constructive Angel Turing machine (pp. 148–152) plus substrate-flexibility; four motivating reasons asserted (pp. 158–159)
C9	Radical solipsism is disconfirmed by the author’s own experiments	moderate (for the author, on his ground rules)	three self-experiments with Fisher/binomial tests, n = 20 each, single subject, 2012 (pp. 173–188)
C10	Under standard cosmology, almost every act causes almost every non-unique future event	moderate as a conditional; the premises are contestable	informal physics argument with explicit assumptions (1)–(4) (p. 206); convergent toy model in n. 265
C11	Consciousness can be defined innocently by example and the concept is non-empty	moderate	definition by example with negative cases; empirical commitment about ordinary concept uptake acknowledged but not tested (p. 232)
C12	Differently structured visual experiences can be equally veridical (Multiple Veridicalities)	weak–moderate	driving-safety literature, introspection, a reading of Stratton and Harris (pp. 239–247)
C13	Whether garden snails are conscious cannot currently be settled; yes, no and gong are all live	moderate	Common Ground Problem and species-specific evidence (pp. 273–281)
C14	AI of debatable personhood is likely and poses a Full Rights Dilemma that credence-weighting does not dissolve	moderate for the dilemma’s structure; weak for the prediction	conceptual argument (pp. 302–306); the likelihood and “soon” rest on technological forecasting and anecdote (teen attitudes, Replika, De Graaf et al. 2021)
C15	Exploring unanswerable questions is among the most intrinsically valuable human activities	assertion	ch. 12, avowedly partisan (“predictable that I would think”, p. 316)

The most citable sentences outrun their support. The PUP overview’s “every possible theory of the relation of mind and cosmos defies common sense” (compare p. 59, “every possible theory”) is an empirical generalisation over the theories actually developed. The body argues it abductively and invites counterexamples; it is not proven over possible theories. “Almost everything you do causes almost everything” holds only under the chapter’s stacked cosmological assumptions (title: “Under Certain Not Wholly Implausible Assumptions”).

Method

A book of informal philosophical argument. Its recurring moves are thought experiments (supersquids, antheads, Angel, Convexia, Robot Alpha), abductive arguments from the history of philosophy, credence-assignment with toy expected-value tables, and trilemma/quadrilemma diagnoses. Chapter 6 alone uses self-experimentation with prespecified predictions and manual Fisher’s exact and binomial tests. It was drafted in the present tense with dummy data from pilot runs, then completed on the day (p. 173).

Concepts

Bizarre — contrary to common sense, where common sense is what most untrained people confidently, if implicitly, believe (p. 65).
Dubious — not epistemically compelling. Wild means bizarre and dubious. A wilderness is a topic where something wild must be among the core truths (pp. 65–66).
Universal Bizarreness / Universal Dubiety — every theory of the mind–world relation is bizarre / none compels belief (p. 86).
Compromise/rejection views — neither materialist, dualist nor idealist: Kant, Russell, Chalmers’s property dualism, “no fact of the matter” (pp. 84–85).
Contiguism — prejudice against spatially discontinuous entities (p. 14).
Anti-nesting principle — a conscious system cannot have conscious parts or be part of a conscious whole (p. 41).
1% skepticism; non-skeptical realism — about a 1% credence that some radically skeptical scenario obtains; the view that none does (pp. 115–116).
Wildcard skepticism — credence reserved for skeptical scenarios one has not considered (p. 115).
Transcendental idealism (as used here) — spatial properties depend on our minds, and things in themselves are unknowable (p. 135).
Empirical object; spatial manifold — objects laid out in one’s immersive spatial environment; an environment in which every part is spatially related to every other (p. 142).
Ripple; puppet; Vast — a chain of successor particles; a far-future counterpart whose act one intentionally causes; a power tower of googolplexes (pp. 204, 211, 207).
Innocence (two kinds) and Wonderfulness — a definition free of dubious metaphysics and epistemology and of dubious scientific theory, which still leaves open the questions people wonder about (p. 219).
Multiple Veridicalities — more than one distinct way an object can appear accurately (p. 240). Teavy/toovy are the pre-inversion and post-inversion looks.
Entity, state, modal, width and property sparseness or abundance — the five dimensions (pp. 268–271).
Common Ground Problem — theories of consciousness rest on prior assumptions about prevalence that opponents reject (§5.1).
Debatable personhood — neither personhood nor falling far short of it is epistemically excluded. Relational and epistemic, not intrinsic (p. 284).
Full Rights Dilemma; Design Policy of the Excluded Middle — see ch. 11 above.
Fission-fusion monster — an entity that divides and merges at will (p. 310).

Connections

LIT-159 is chapter 2 of the manuscript and chapter 3 of the published book. The book version is longer. It moves the anti-nesting critique into an appendix, adds the reply to Kammerer, greatly expands the IIT footnote and replaces the paper’s three-way hedge with a four-way choice that rejects quietism (see corrections). Its role changes too. The paper was a stand-alone challenge; in the book it is the lead case study for Universal Bizarreness (ch. 3 §3.4).

LIT-191 (Schwitzgebel, AI and Consciousness, 2025–26) is the later, book-length continuation of chapters 10 and 11. The present book already has the Common Ground Problem, the species-specificity of introspective evidence, debatable personhood and the Design Policy of the Excluded Middle. Where LIT-191 develops the Mimicry Argument and the Leapfrog Hypothesis (per its record), this book’s AI material is still tied to GPT-3 and a hypothetical embodied “GPT-6”. The published version updates this to ChatGPT (index, pub. pp. 231–32).

LIT-166 (Schwitzgebel & Sinnott-Armstrong 2026, a critical review of Birch, Sebo and Keane) takes up the ethical side of chapter 11. Chapter 11 considers credence-weighted rights (“15% of a human vote”) and finds that they “only partly mitigate” the dilemma (p. 306). LIT-166, per its summary, presses the radical conditional that precautionary principles could let aggregate insect or AI interests outweigh humanity’s. The book cites Birch (forthcoming) on bees and “theory-light” methods (ch. 10 nn. 343–344). LIT-166 is still Deferred (not read), so this link comes from its summary only.

LIT-096 (Nagel 1974) supplies the book’s working synonym. “Something it’s like” is one of the equivalent phrases chapter 8 accepts, though n. 278 notes it “might invite the thought that there is one typical thing it is like”. Chapter 10 opens with Nagel’s phrasing and cites Nagel for the difficulty of extrapolating beyond the human case (n. 348). The book adopts Nagel’s criterion but not his anti-reductionist conclusion. It keeps materialism as its 55% favourite.

LIT-185 (Barrett et al. 2026 on IIT) bears directly on the book’s two IIT arguments. The first is that the Exclusion Postulate generates one-ballot and sleeping-citizen consciousness flips, with an Aaronson expander-graph construction for arbitrarily large Φ (ch. 2 appendix, n. 56). The second is that IIT is “very nearly panpsychist”, so a reader drawn to sparse views can reasonably treat it as “a theory of something other than conscious experience” (ch. 10, pp. 275–276). LIT-185 defends IIT against exactly the expander-grid and panpsychism charges, but through the authors’ reinterpretation, not IIT 4.0 as stated (per its status note). The two should be read together.

LIT-056 (Butlin et al. 2023, indicator properties for AI consciousness) appeared after this manuscript. It is the kind of “theory-light” or indicator approach that chapter 10 argues must still rest on contestable assumptions about prevalence (pp. 277–278). Whether the published book discusses it is unverified.

Account of consciousness held. The book holds no theory of consciousness. It holds three things. (i) Realism about phenomenal consciousness in an “innocent” sense defined by example (ch. 8), which is explicitly anti-eliminativist. (ii) Agnosticism about its distribution and metaphysics, leaning materialist (about 55%). (iii) The meta-thesis that every theory of consciousness is bizarre and dubious and that the prevalence question is currently unsettleable (the Common Ground Problem). The consciousness tag is justified. Four of eleven substantive chapters (2, 8, 10, 11) and much of chapter 3 are about phenomenal consciousness, its theories and its attribution to groups, animals and machines, which matches the tag’s blurb exactly.

Bearing on the record

It supports treating the record’s consciousness-attribution works as operating in a “wilderness”. Any indicator-based verdict (compare LIT-056) inherits the Common Ground Problem.
For LIT-159: the book version supersedes the paper as the fuller statement (the Kammerer reply, expanded IIT critique and quietism rejected). The paper remains the citable peer-reviewed version. No status change is suggested for LIT-159. A relates or extends link from this book to LIT-159 would be accurate.
ML practice: nothing here for training or evaluation practice in the Anthology of the SOTA. Chapter 11’s AI material is about moral status and design policy, and its technical picture (GPT-3, 175B parameters, 96 layers, p. 285) is illustrative and dated. The Design Policy of the Excluded Middle is a policy recommendation about which AI systems to build, not an ML method.

Limitations

Version. The text read is the June 2022 final manuscript. The published book (Jan 2024) reorders chapters, restructures the objections appendix, retitles ch. 8, updates ch. 11 for ChatGPT and revises ch. 1. Quotations and page numbers here are from the manuscript and may not match the published book.
Support. Apart from chapter 6, every result is an informal argument or an illustrative credence calculation. The numbers (1%, 55%, 15%, 99.8%) are the author’s judgments, and he says they are not measurements (p. 116).
Universal Bizarreness is inductive. It rests on the author’s reading of the history of philosophy and a standing challenge to produce a counterexample. It is an empirical generalisation over developed theories, not a result about all possible theories.
Chapter 6’s evidence is from a single subject, n = 20 per test, run in 2012 by the interested party. The author notes the anti-solipsism is thin: it establishes “almost nothing about the metaphysical character of the external world” (p. 190). Co-author Moore’s postscript notes it does not work from his side (p. 191).
Chapter 7 stacks contestable cosmological premises: infinitude, post-heat-death fluctuations at a non-decreasing rate, non-terminating ripples. It says so. The conclusion is conditional on them.
Chapter 11’s forecasts (“soon”, and the teen-attitudes evidence) are anecdote plus one survey (De Graaf et al. 2021). They were written before ChatGPT.
Draft defects (unresolved “note XXX” cross-references, broken section numbering, references missing several source articles) are listed under corrections.

Open questions

Is there any well-developed metaphysics of mind that violates no common-sense commitment? That would refute C1. The author issues this challenge (p. 63).
What would “in-between” or borderline consciousness be? Chapter 10 says there is “no well-developed theory” of it (p. 267), and the author’s 2022 manuscript on borderline consciousness is cited as work in progress.
Could any criterion of consciousness be independently motivated, exclude the US and still include rabbits and supersquids? The book treats this as the materialist’s open “challenge” (p. 40, appendix §6).
Is there a non-question-begging way past the Common Ground Problem, for example convergent evidence across rival theories? The book leaves this to the long run (p. 281).
How does the published text of chapters 2/3 and 11 differ from the manuscript? Reading the published book, or at least the published chapter 3, the appendix and chapter 11, would settle this.

Corrections to the seeded skim

No dossier existed for this work. It was seeded from the request alone, so there is no skim to correct. The corrections below are to the request’s framing and to what the record assumes.
LIT-159 is a chapter, but not chapter 2 of the published book. In the manuscript it is chapter 2 (pp. 11–40), with a 17-page “Chapter Two Appendix: Six Objections” (pp. 41–57). The published table of contents (pub. pp. vii–viii) reorders Part I. Chapter 2 becomes “Universal Bizarreness and Universal Dubiety” and the US argument becomes chapter 3 (pub. pp. 47–71), in a new Part II, “Peculiar Possibilities”. The appendix moves to the end of the book as “Appendix: Five More Objections to U.S. Consciousness” (pub. pp. 261–275). The published index places anti-nesting (261–65), speed/synchrony (265–66), Dennett (266–67), Chalmers (268–69) and representation (270–72) there. That is five objections. The manuscript’s sixth, “Methodological Concerns”, is either folded into the chapter body or dropped. Which one is unverified.
How the book’s US chapter differs from the paper read as LIT-159 (NOTE-131, which read the 2014 manuscript). (i) The paper’s §2 critique of anti-nesting (Putnam; Tononi’s Exclusion Postulate) leaves the body and becomes Appendix Objection 1. (ii) A new objection is answered: Kammerer (2015), paired with Dretske as Objection 5 (pp. 52–55). His principle (LIT-409, NOTE-362) has two conditions and applies state by state. A whole lacks a given conscious state if (A) some part plays a role that requires it to consciously represent the whole, and (B) without any part playing such a role, the whole would lose the properties sufficient for that state. This note glossed it at v1 as “a whole is not conscious if its conscious parts represent the whole”, which drops condition B and the state-by-state scope. Schwitzgebel calls it “motivated post-hoc” and gives it the same counter-consequence as Dretske’s: you “would lose consciousness upon inhaling a Planck-sized person” (p. 54). (iii) The IIT critique is much expanded in n. 56. It takes on Tononi & Koch 2015’s claim that the US must have lower Φ than its citizens and cites Hoel, Albantakis & Tononi 2013 in the US’s favour. It adds List 2018 and the “Max goes to sleep” flicker case: when the highest-Φ citizen falls asleep, every person in the system loses consciousness under Exclusion. It also adds an Aaronson expander-graph construction for arbitrarily large Φ. (iv) The paper ended undecided among three readings. The book ends with four choices (p. 39–40): accept US consciousness, reject materialism, treat the argument as a challenge to materialists, or quietism. It rejects quietism (“Oh, how horrible that fourth reaction is!”) and prefers “some indecisive superposition” of the other three. (v) The Exxon-Mobil example becomes Google and an aircraft carrier (p. 39). (vi) Eliminativism is sharpened with Frankish 2016 and Kammerer 2021, and a “quasi-consciousness” escape is noted (Lee 2019, n. 37). (vii) The “un-Copernican” point is grounded in the author’s “Copernican Principle of consciousness” (2020b, n. 38). (viii) A manuscript slip: n. 42 speaks of “four prominent … theories” but names three (Humphrey’s, Dennett’s, Tononi’s pre-2012 view). The paper’s list also included Dretske.
The manuscript is a draft with visible scaffolding. It has unresolved “note XXX” cross-references (pp. 37, 50, 89, 158, 276). Section numbers jump: ch. 5 goes from §6 to §9 (p. 159), and ch. 10 has “4.2. No.” between §§3.1 and 3.3 (p. 262) and “5..” and “6..” headings. It has production notes for illustrations, an acknowledgements page reading ”*** Your name could be here. ***” (p. 320), and stale cross-references (ch. 3 n. 77 “note 5 above”, n. 104 “note 16 above”). The references omit the source articles of chs 6, 7, 9, 10 and 11. None of this affects the arguments, but page numbers and wording will not match the published text.
The published book changes more than order. Chapter 8 is retitled “Consciousness, Innocent and Wonderful” (manuscript: “An Innocent and Wonderful Definition of Consciousness”). Chapter 11 loses its subtitle “Doubts and a Dilemma”. The published index has “large language model programs (ChatGPT, etc.), 231–32, 303n5”, while the manuscript’s chapter 11 discusses only GPT-3 and a hypothetical “GPT-6 mall cop”. “Standard-Issue Naturalism” (manuscript p. 316) appears in the published index as “Mainstream Naturalism”. The far-future counterparts of chapter 7 (“you-prime”) appear in the published index as “Bizarros”. Published chapter 1 drops the section “To Non-Specialists: An Invitation and Apology” and adds “Each chapter is meant to be separately comprehensible” (pub. p. 7).
Chapter sources, verified on Crossref. Ch. 2: Philosophical Studies 172 (2015), 1697–1721 (LIT-159). Ch. 3: “The Crazyist Metaphysics of Mind”, Australasian J. Phil. 2014, doi 10.1080/00048402.2014.910675. The manuscript renames “crazy/crazyism” to “wild” because of the term’s “ableist history” (n. 83). Ch. 4: Noûs 51 (2017). Ch. 5: Disputatio 55 (2019). Ch. 6: Schwitzgebel & Moore, J. Am. Phil. Assoc. 2015, doi 10.1017/apa.2014.27. Ch. 8: J. Consciousness Studies 23 (2016). Ch. 10: Philosophical Topics 2020, doi 10.5840/philtopics20204813. Ch. 11: related to “The Full Rights Dilemma for AI Systems of Debatable Moral Personhood” (Robonomics 2023) and to Schwitzgebel & Garza 2015/2020. Chs 7 and 9: no separate source article was found.
primary topic: metaphysics (if the record’s convention for this batch wants consciousness first, that is defensible too; see Connections).

nucleation

Explorer