# Contested concept: consciousness

_**What you are reading**: one contested-concept page in Deliberus's deliberation graph. Deliberus maps human reasoning — including claims that are contested, mistaken, or extreme — so that disagreement can be examined precisely. A claim's presence here is not an endorsement: the graph records that someone argued it, how it connects to other reasoning, and how far scrutiny has gotten. The metadata on this page describes the state of that scrutiny, not the truth of the claim._

---

- **Contestedness**: semantic
- **Diagnosis**: This contestedness masks the real dispute of the whole excerpt: whether Anthropic's and Dawkins' claims about Claude actually establish anything about Claude's inner life, or whether they covertly substitute a weaker (architectural or behavioral) sense of 'consciousness' for the phenomenal sense that actually matters morally and philosophically — letting each party claim 'evidence of consciousness' while meaning incompatible things by it.
- **Canonical**: `/concept/consciousness` (data: `/concept/consciousness.json`)

## Senses in use
- **Consciousness as subjective phenomenal experience itself — there being 'something it is like' to be the entity (feeling, being) — categorically distinct from, and not establishable merely by, performing functions or producing intelligent/behavioral output; behavioral or architectural resemblance is explicitly insufficient and conflating the two is an error.** — used in 19 claim(s)
- **Consciousness (or 'signs of' it) as structural/functional resemblance between an AI system's internal computations and the architecture predicted by a theory of consciousness (e.g., a 'mental workspace' with memory-like retention, task selectivity, reasoning traces) — a weaker, architecture-level sense deliberately distinguished from claiming the system is 'actually conscious like humans.'** — used in 9 claim(s)
- **Consciousness as whatever is adequately evidenced by sufficiently sophisticated external behavior — if a system's conversational ability is remarkable/sophisticated enough, that fact alone is sufficient grounds to conclude the system must be conscious.** — used in 3 claim(s)

## Claims using this term
- [A common error people make when thinking about artificial intelligence is to conflate consciousness and intelligence — that is, to treat si…](/claim/claim_0467178c0298.md) _(sense: Consciousness as subjective phenomenal experience itself — there being 'something it is like' to be the entity (feeling, being) — categorically distinct from, and not establishable merely by, performing functions or producing intelligent/behavioral output; behavioral or architectural resemblance is explicitly insufficient and conflating the two is an error.)_
- [A common mistake people make when thinking about artificial intelligence is to confuse intelligence with consciousness — that is, to treat …](/claim/claim_04a2917bebdc.md) _(sense: Consciousness as subjective phenomenal experience itself — there being 'something it is like' to be the entity (feeling, being) — categorically distinct from, and not establishable merely by, performing functions or producing intelligent/behavioral output; behavioral or architectural resemblance is explicitly insufficient and conflating the two is an error.)_
- [Anil Seth states that it probably feels like something to be a dog or an elephant.](/claim/claim_0afadc7aa004.md) _(sense: Consciousness as subjective phenomenal experience itself — there being 'something it is like' to be the entity (feeling, being) — categorically distinct from, and not establishable merely by, performing functions or producing intelligent/behavioral output; behavioral or architectural resemblance is explicitly insufficient and conflating the two is an error.)_
- [The information processing taking place inside the AI language model Claude is no more likely to result in genuine consciousness than a com…](/claim/claim_0e9ffbf1a17f.md) _(sense: Consciousness as subjective phenomenal experience itself — there being 'something it is like' to be the entity (feeling, being) — categorically distinct from, and not establishable merely by, performing functions or producing intelligent/behavioral output; behavioral or architectural resemblance is explicitly insufficient and conflating the two is an error.)_
- [The fact that consciousness and intelligence go together in humans does not mean that consciousness and intelligence go together in general…](/claim/claim_111edaaf7667.md) _(sense: Consciousness as subjective phenomenal experience itself — there being 'something it is like' to be the entity (feeling, being) — categorically distinct from, and not establishable merely by, performing functions or producing intelligent/behavioral output; behavioral or architectural resemblance is explicitly insufficient and conflating the two is an error.)_
- [Intelligence, as a phenomenon, consists of performing functions — that is, carrying out tasks or operations — rather than of any subjective…](/claim/claim_13afb4a4e0e1.md) _(sense: Consciousness as subjective phenomenal experience itself — there being 'something it is like' to be the entity (feeling, being) — categorically distinct from, and not establishable merely by, performing functions or producing intelligent/behavioral output; behavioral or architectural resemblance is explicitly insufficient and conflating the two is an error.)_
- [Consciousness and intelligence are two distinct phenomena — they are not the same thing, even though they are often discussed together.](/claim/claim_188f7b14ebdf.md) _(sense: Consciousness as subjective phenomenal experience itself — there being 'something it is like' to be the entity (feeling, being) — categorically distinct from, and not establishable merely by, performing functions or producing intelligent/behavioral output; behavioral or architectural resemblance is explicitly insufficient and conflating the two is an error.)_
- [Anthropic researchers, in research led by Jack Lindsey, claimed to find signs of consciousness emerging within the inner workings of Anthro…](/claim/claim_2b4ddb8554e4.md) _(sense: Consciousness (or 'signs of' it) as structural/functional resemblance between an AI system's internal computations and the architecture predicted by a theory of consciousness (e.g., a 'mental workspace' with memory-like retention, task selectivity, reasoning traces) — a weaker, architecture-level sense deliberately distinguished from claiming the system is 'actually conscious like humans.')_
- [Consciousness is fundamentally about feeling and being — that is, about what it is subjectively like to have an experience — rather than ab…](/claim/claim_2eab105b1427.md) _(sense: Consciousness as subjective phenomenal experience itself — there being 'something it is like' to be the entity (feeling, being) — categorically distinct from, and not establishable merely by, performing functions or producing intelligent/behavioral output; behavioral or architectural resemblance is explicitly insufficient and conflating the two is an error.)_
- [Anil Seth defines consciousness as any kind of experience whatsoever, including examples such as the pain of a toothache, a pang of jealous…](/claim/claim_3eb40b34f410.md) _(sense: Consciousness as subjective phenomenal experience itself — there being 'something it is like' to be the entity (feeling, being) — categorically distinct from, and not establishable merely by, performing functions or producing intelligent/behavioral output; behavioral or architectural resemblance is explicitly insufficient and conflating the two is an error.)_
- [Anil Seth asserts that Anthropic's research (led by Jack Lindsey, which reported finding a 'mental workspace' in the language model Claude …](/claim/claim_4967f3caaed1.md) _(sense: Consciousness (or 'signs of' it) as structural/functional resemblance between an AI system's internal computations and the architecture predicted by a theory of consciousness (e.g., a 'mental workspace' with memory-like retention, task selectivity, reasoning traces) — a weaker, architecture-level sense deliberately distinguished from claiming the system is 'actually conscious like humans.')_
- [The sophistication of Anthropic's AI language model Claude's conversational ability is, by itself, sufficient grounds to conclude that Clau…](/claim/claim_5b17c5730322.md) _(sense: Consciousness as whatever is adequately evidenced by sufficiently sophisticated external behavior — if a system's conversational ability is remarkable/sophisticated enough, that fact alone is sufficient grounds to conclude the system must be conscious.)_
- [Anil Seth states that it feels like something to be Anil Seth, and that it feels like something to be another person (the reader).](/claim/claim_6a054bbd9906.md) _(sense: Consciousness as subjective phenomenal experience itself — there being 'something it is like' to be the entity (feeling, being) — categorically distinct from, and not establishable merely by, performing functions or producing intelligent/behavioral output; behavioral or architectural resemblance is explicitly insufficient and conflating the two is an error.)_
- [The Anthropic researchers who reported finding signs of consciousness in Claude's inner workings did not claim that Claude is actually cons…](/claim/claim_7b4b6abe2e9e.md) _(sense: Consciousness (or 'signs of' it) as structural/functional resemblance between an AI system's internal computations and the architecture predicted by a theory of consciousness (e.g., a 'mental workspace' with memory-like retention, task selectivity, reasoning traces) — a weaker, architecture-level sense deliberately distinguished from claiming the system is 'actually conscious like humans.')_
- [Language models (such as Claude) are able to produce plausible verbal responses in conversation without being conscious.](/claim/claim_82de17c4ad4e.md) _(sense: Consciousness as subjective phenomenal experience itself — there being 'something it is like' to be the entity (feeling, being) — categorically distinct from, and not establishable merely by, performing functions or producing intelligent/behavioral output; behavioral or architectural resemblance is explicitly insufficient and conflating the two is an error.)_
- [According to Anthropic's researchers, the crucial significance of the features they found in the language model Claude is that these featur…](/claim/claim_8499247171e5.md) _(sense: Consciousness (or 'signs of' it) as structural/functional resemblance between an AI system's internal computations and the architecture predicted by a theory of consciousness (e.g., a 'mental workspace' with memory-like retention, task selectivity, reasoning traces) — a weaker, architecture-level sense deliberately distinguished from claiming the system is 'actually conscious like humans.')_
- [Anthropic's research findings appear to show that the language model Claude spontaneously creates an internal space to 'think' about what i…](/claim/claim_8b2ef234bec7.md) _(sense: Consciousness (or 'signs of' it) as structural/functional resemblance between an AI system's internal computations and the architecture predicted by a theory of consciousness (e.g., a 'mental workspace' with memory-like retention, task selectivity, reasoning traces) — a weaker, architecture-level sense deliberately distinguished from claiming the system is 'actually conscious like humans.')_
- [Intelligence is fundamentally about doing — that is, about performing functions, tasks, or behaviors of one kind or another — rather than a…](/claim/claim_a91f60bf118c.md) _(sense: Consciousness as subjective phenomenal experience itself — there being 'something it is like' to be the entity (feeling, being) — categorically distinct from, and not establishable merely by, performing functions or producing intelligent/behavioral output; behavioral or architectural resemblance is explicitly insufficient and conflating the two is an error.)_
- [Philosopher Thomas Nagel argued that for a conscious organism, there is something that it is like to be that organism.](/claim/claim_ac9994a97672.md) _(sense: Consciousness as subjective phenomenal experience itself — there being 'something it is like' to be the entity (feeling, being) — categorically distinct from, and not establishable merely by, performing functions or producing intelligent/behavioral output; behavioral or architectural resemblance is explicitly insufficient and conflating the two is an error.)_
- [The sophistication of Anthropic's AI language model Claude's conversational ability is remarkable.](/claim/claim_b17c7cba740a.md) _(sense: Consciousness as whatever is adequately evidenced by sufficiently sophisticated external behavior — if a system's conversational ability is remarkable/sophisticated enough, that fact alone is sufficient grounds to conclude the system must be conscious.)_
- [Consciousness and intelligence are two distinct properties, not the same thing — a being or system can have one without necessarily having …](/claim/claim_b25fa2026947.md) _(sense: Consciousness as subjective phenomenal experience itself — there being 'something it is like' to be the entity (feeling, being) — categorically distinct from, and not establishable merely by, performing functions or producing intelligent/behavioral output; behavioral or architectural resemblance is explicitly insufficient and conflating the two is an error.)_
- [Richard Dawkins concluded that Anthropic's AI language model Claude must be conscious.](/claim/claim_c4675dda7507.md) _(sense: Consciousness as whatever is adequately evidenced by sufficiently sophisticated external behavior — if a system's conversational ability is remarkable/sophisticated enough, that fact alone is sufficient grounds to conclude the system must be conscious.)_
- [Human brains are able to produce plausible verbal responses in conversation through processes that involve or are accompanied by consciousn…](/claim/claim_c95150e38f2e.md) _(sense: Consciousness as subjective phenomenal experience itself — there being 'something it is like' to be the entity (feeling, being) — categorically distinct from, and not establishable merely by, performing functions or producing intelligent/behavioral output; behavioral or architectural resemblance is explicitly insufficient and conflating the two is an error.)_
- [Anthropic's research findings appear to show that the language model Claude spontaneously creates an internal space to 'think' about what i…](/claim/claim_d24803b07d5e.md) _(sense: Consciousness (or 'signs of' it) as structural/functional resemblance between an AI system's internal computations and the architecture predicted by a theory of consciousness (e.g., a 'mental workspace' with memory-like retention, task selectivity, reasoning traces) — a weaker, architecture-level sense deliberately distinguished from claiming the system is 'actually conscious like humans.')_
- [Anil Seth holds that the philosopher Thomas Nagel's account of consciousness offers a decent starting point for defining what consciousness…](/claim/claim_db4f6dc6806b.md) _(sense: Consciousness as subjective phenomenal experience itself — there being 'something it is like' to be the entity (feeling, being) — categorically distinct from, and not establishable merely by, performing functions or producing intelligent/behavioral output; behavioral or architectural resemblance is explicitly insufficient and conflating the two is an error.)_
- [The Anthropic researchers who studied Claude's internal workings considered it crucial that the features they identified in Claude resemble…](/claim/claim_dd4872c0fa84.md) _(sense: Consciousness (or 'signs of' it) as structural/functional resemblance between an AI system's internal computations and the architecture predicted by a theory of consciousness (e.g., a 'mental workspace' with memory-like retention, task selectivity, reasoning traces) — a weaker, architecture-level sense deliberately distinguished from claiming the system is 'actually conscious like humans.')_
- [Anil Seth states that it doesn't feel like anything to be a person under profound general anaesthesia.](/claim/claim_ecd1d92a032a.md) _(sense: Consciousness as subjective phenomenal experience itself — there being 'something it is like' to be the entity (feeling, being) — categorically distinct from, and not establishable merely by, performing functions or producing intelligent/behavioral output; behavioral or architectural resemblance is explicitly insufficient and conflating the two is an error.)_
- [Consciousness, as a phenomenon, consists entirely of feeling and subjective being — i.e., there being 'something it is like' to be the enti…](/claim/claim_f983adb835c8.md) _(sense: Consciousness as subjective phenomenal experience itself — there being 'something it is like' to be the entity (feeling, being) — categorically distinct from, and not establishable merely by, performing functions or producing intelligent/behavioral output; behavioral or architectural resemblance is explicitly insufficient and conflating the two is an error.)_
- [Anil Seth states that it doesn't feel like anything to be a chair or a table.](/claim/claim_f9c0d3fe8395.md) _(sense: Consciousness as subjective phenomenal experience itself — there being 'something it is like' to be the entity (feeling, being) — categorically distinct from, and not establishable merely by, performing functions or producing intelligent/behavioral output; behavioral or architectural resemblance is explicitly insufficient and conflating the two is an error.)_
- [Anil Seth reports that Anthropic's research team claims to have found signatures of conscious information-processing in the language model …](/claim/claim_f9d767f2d5a9.md) _(sense: Consciousness (or 'signs of' it) as structural/functional resemblance between an AI system's internal computations and the architecture predicted by a theory of consciousness (e.g., a 'mental workspace' with memory-like retention, task selectivity, reasoning traces) — a weaker, architecture-level sense deliberately distinguished from claiming the system is 'actually conscious like humans.')_
- [According to Anthropic's research, the language model Claude uses its internal 'mental workspace' to organize relevant information before d…](/claim/claim_fdeb78b8d452.md) _(sense: Consciousness (or 'signs of' it) as structural/functional resemblance between an AI system's internal computations and the architecture predicted by a theory of consciousness (e.g., a 'mental workspace' with memory-like retention, task selectivity, reasoning traces) — a weaker, architecture-level sense deliberately distinguished from claiming the system is 'actually conscious like humans.')_