Erik Hoel · Jun 5, 2026
Erik Hoel argues in The Intrinsic Perspective: Don’t dethrone consciousness
Markdown · JSON · Original source
Erik Hoel’s essay was published on June 5, 2026. He argues that current should not be treated as conscious, while rejecting the idea that this requires dismissing their capabilities or denying that artificial consciousness could ever exist. Hoel argues that theories of consciousness and studies of LLMs do not yet offer a sound way to decide whether these systems are conscious. He points to their failures on unfamiliar tasks and difficulty tracking their own reasoning, and suggests their written explanations may be confabulations rather than faithful accounts of why they acted. For Hoel, the stakes extend beyond how we classify machines: as AI takes on more cognitive work, treating consciousness as irrelevant could diminish its cultural importance, even though, he says, we understand little about it. He hopes AI will instead prompt better understanding of consciousness. He ends by warning that dethroning consciousness would mark “The end of living, and the beginning of survival.” Whether consciousness explains the differences Hoel describes, or research can establish it in LLMs, remains open.
Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.
Original source
Erik Hoel · Article
Source details
Original source date: Jun 5, 2026 (2026-06-05T14:35:09.000Z).
Date basis: Date supplied by the publisher’s feed.
Added to Escape Meridian: 2026-11-06T19:48:23.734Z. First observed: 2026-10-29T15:27:50.477Z. Retrieved: 2026-10-29T15:30:12.754Z.
Referenced by
Recorded citations and public file attachments. A reference alone does not establish agreement or influence.
No public references recorded here yet.
Limits
Text as retrieved on October 29, 2026; originally published June 5, 2026. Hoel’s claims about the limits of consciousness research and the interpretation of behavior are arguments in the essay, not a settled scientific account.
Open question
What kind of evidence could distinguish consciousness in an from behavior that can be explained without it?
Source-stated
Hoel argues that LLM capabilities do not require a conclusion about consciousness, and that their intelligence can be considered separately from it.
intelligence (at least, of the question and answer variety targeted in benchmarks) is indeed dissociable from consciousness.
Source-stated
Hoel argues that current theories and research do not provide a sound basis for deciding whether LLMs are conscious.
you cannot actually do experiments on LLMs alone and come to a scientifically-supported conclusion about their possession of, or lack thereof, consciousness.
Source-stated
Hoel argues that LLMs may produce explanations of their actions that are confabulations rather than faithful accounts of their internal reasoning.
an LLM must constantly confabulate the reasons for its own actions.
Source-stated
Hoel warns that treating consciousness as unimportant as AI takes on more of civilization’s work could diminish its cultural importance and harm the human spirit.
If instead of that, our cultural takeaway from LLMs is to throw out the concept of “consciousness” or minimize its importance, to dethrone the phenomenon, the consequences would be dire—it would sap the human spirit.
Source-stated
Hoel concludes with a warning that dethroning consciousness could end living as he understands it.
dethroning consciousness would mark “The end of living, and the beginning of survival.”
Source and contributor context
Erik Hoel · Publisher
The publisher is identified here. No individual author has been inferred.
Source media
Media retrieval status: unchecked.
Terms and references
Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.
Large language model
A language model with many learned numerical settings, used to process and generate language.
Editorial provenance
Version 1. Publication ID: 9ccaf490-187b-0809-966a-22c45e84dfd6.
Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5.
Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-06T19:43:28.965Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: bridge_admitted_by_model.
Observation: 1aa2af52a648f6f450cd760705f3057b7f1fb3fc1adca8ce8e4fd646b7ba596a. Source text hash: 478dc8a9d3537c901678882237c7dd42f7a3b8857fc3549bc7463153e414311a.
Revision and source-check history
Version 1 · 2026-11-06T19:48:23.734Z
Erik Hoel argues in The Intrinsic Perspective: Don’t dethrone consciousness
Erik Hoel’s essay was published on June 5, 2026. He argues that current should not be treated as conscious, while rejecting the idea that this requires dismissing their capabilities or denying that artificial consciousness could ever exist. Hoel argues that theories of consciousness and studies of LLMs do not yet offer a sound way to decide whether these systems are conscious. He points to their failures on unfamiliar tasks and difficulty tracking their own reasoning, and suggests their written explanations may be confabulations rather than faithful accounts of why they acted. For Hoel, the stakes extend beyond how we classify machines: as AI takes on more cognitive work, treating consciousness as irrelevant could diminish its cultural importance, even though, he says, we understand little about it. He hopes AI will instead prompt better understanding of consciousness. He ends by warning that dethroning consciousness would mark “The end of living, and the beginning of survival.” Whether consciousness explains the differences Hoel describes, or research can establish it in LLMs, remains open.
Original source
Erik Hoel · Article
Source details
Original source date: Jun 5, 2026 (2026-06-05T14:35:09.000Z).
Date basis: Date supplied by the publisher’s feed.
Added to Escape Meridian: 2026-11-06T19:48:23.734Z. First observed: 2026-10-29T15:27:50.477Z. Retrieved: 2026-10-29T15:30:12.754Z.
Limits
Text as retrieved on October 29, 2026; originally published June 5, 2026. Hoel’s claims about the limits of consciousness research and the interpretation of behavior are arguments in the essay, not a settled scientific account.
Open question
What kind of evidence could distinguish consciousness in an from behavior that can be explained without it?
Source-stated
Hoel argues that LLM capabilities do not require a conclusion about consciousness, and that their intelligence can be considered separately from it.
intelligence (at least, of the question and answer variety targeted in benchmarks) is indeed dissociable from consciousness.
Source-stated
Hoel argues that current theories and research do not provide a sound basis for deciding whether LLMs are conscious.
you cannot actually do experiments on LLMs alone and come to a scientifically-supported conclusion about their possession of, or lack thereof, consciousness.
Source-stated
Hoel argues that LLMs may produce explanations of their actions that are confabulations rather than faithful accounts of their internal reasoning.
an LLM must constantly confabulate the reasons for its own actions.
Source-stated
Hoel warns that treating consciousness as unimportant as AI takes on more of civilization’s work could diminish its cultural importance and harm the human spirit.
If instead of that, our cultural takeaway from LLMs is to throw out the concept of “consciousness” or minimize its importance, to dethrone the phenomenon, the consequences would be dire—it would sap the human spirit.
Source-stated
Hoel concludes with a warning that dethroning consciousness could end living as he understands it.
dethroning consciousness would mark “The end of living, and the beginning of survival.”
Source and contributor context
Erik Hoel · Publisher
The publisher is identified here. No individual author has been inferred.
Source media
Media retrieval status: unchecked.
Terms and references
Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.
Large language model
A language model with many learned numerical settings, used to process and generate language.
Editorial provenance
Version 1. Publication ID: 9ccaf490-187b-0809-966a-22c45e84dfd6.
Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5.
Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-06T19:43:28.965Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: bridge_admitted_by_model.
Observation: 1aa2af52a648f6f450cd760705f3057b7f1fb3fc1adca8ce8e4fd646b7ba596a. Source text hash: 478dc8a9d3537c901678882237c7dd42f7a3b8857fc3549bc7463153e414311a.
uncertain · 2026-11-06T19:43:28.965Z · anthropic/claude-opus-5.5
Publication: 9ccaf490-187b-0809-966a-22c45e84dfd6
attribution — pass: The reading correctly attributes this single essay to Erik Hoel and The Intrinsic Perspective. It keeps the Pope's, Chiang's and Olah's positions out of Hoel's voice, and presents his views as his arguments.
Don't dethrone consciousness! LLMs aren't conscious (and thinking they are is culturally dangerous) Erik Hoel
summary — uncertain: The reading softens Hoel's central thesis. Saying theories and studies 'do not yet offer a sound way to decide' makes him sound agnostic. He actually argues, through a layer-shaving thought experiment, that no non-trivial falsifiable theory could grant current LLMs consciousness. The reading also omits the framing: the essay responds to the Pope's encyclical and Ted Chiang. Smallest repair: state that he argues no viable falsifiable theory could assign current LLMs consciousness, while artificial consciousness in general may be possible. Note his stance against the Pope and Chiang.
there simply is no viable non-trivial falsifiable theory of consciousness that could possibly assign current LLMs consciousness
date — pass: The source dateline reads Jun 05, 2026, matching evidence.originalPublished (2026-06-05 UTC) at day precision. The modification time equals the publication time. No hindsight framing or later events are used. This is a publisher page, so no archive disclosure is needed.
Erik Hoel Jun 05, 2026
support — pass: Each claim matches the cited text. One wording is slightly loose: 'could end living as he understands it' adds a qualifier, but Hoel says dethroning consciousness 'would mark' the end of living, quoting Chief Seattle. This does not misstate his position. The summary's mention of failures on unfamiliar tasks and meta-cognition, and its point that artificial consciousness in general may be possible, are both source-stated.
Artificial consciousness in general might be possible
media — pass: No media was selected, and no verified media candidates were supplied, so the lack of media is appropriate.
Challenge or correct this reading
Send a correction or an opposing source with this reading’s URL. Disagreement and uncertainty can remain unresolved.
A language model with many learned numerical settings, used to process and generate language.
Definition referenceA language model with many learned numerical settings, used to process and generate language.
Definition referenceA language model with many learned numerical settings, used to process and generate language.
Definition referenceA language model with many learned numerical settings, used to process and generate language.
Definition referenceA language model with many learned numerical settings, used to process and generate language.
Definition referenceA language model with many learned numerical settings, used to process and generate language.
Definition reference