Pith. sign in

REVIEW 4 major objections 4 minor 79 references

Some hypotheses on how chatbots work in problem-solving-driven conversations. Large Language Models as confirmation of the Innovation Illusion

T0 review · 4 major / 4 minor · reviewed 2026-08-04 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A chatbot, as a conversation partner, is not and cannot become an analytical thinking partner through text-only training.

desk verdict A clearly written speculative essay that is honest about its own status, but the central impossibility claim rests on an unmeasured 'onward text' premise and an untested reconstruction hypothesis; worth a serious discussion, not a conclusive result. read the letter →

arxiv 2606.07722 v4 pith:ZZX3ATEB submitted 2026-06-05 cs.AI

S.F.M. van Vlijmen , H.D. Lethe jr This is my paper
classification cs.AI
keywords chatbotslargelanguagemodelsmetaphoricalproblempropagationonwardtextSystem1and2thinkinganalogicalreasoningpredictiveprocessingproblem-solvingconversations
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper argues that chatbots, built on large language models, are fundamentally System 1-style conversation partners: fluent, fast, and problem-to-solution oriented, but incapable of the reflective, analytical thinking that defines System 2 reasoning. It introduces a model of human thought called 'metaphorical problem propagation' — networks of problem positions and solution steps shaped by metaphor and prediction — and hypothesizes that LLM training reconstructs reduced, 'onward' forms of these from text. Because most training texts are solution-oriented, non-reflective prose, and because the resulting model lacks an embodied, experience-based conceptual system, the chatbot cannot assess or correct its own outputs the way a human thinking partner can. The conclusion is that further development of LLMs alone will not turn chatbots into analytical thinking partners; their role must be understood as guided cognitive assistance, not independent analysis.

What carries the argument

The central object is the 'metaphorical problem propagation' — a model of the human thought space as a network of problem positions and solution steps, structured by metaphors and powered by predictive processing. It carries the argument by linking human cognition, human text, and LLM behavior: thought is modeled as propagation through this space; text is a reduced verbalization of it; LLM training is said to reconstruct artificial versions of it from text. The 'onward text' hypothesis is the load-bearing premise: most training text is assumed to be solution-oriented, coherent, non-experimental prose that moves from problem to conclusion without reflection, and this determines the System 1-l

What would settle it

Compile a training corpus of deliberately analytical, reflective texts — for example, philosophical debates, contradictory position papers, and open-ended problem analyses that withhold conclusions — train an LLM on it, and test whether it reliably catches its own inconsistencies, revises assumptions on feedback, and withholds answers under uncertainty. If it does so robustly, the paper's claim that text-only training cannot produce analytical thinking would be falsified.

Watch

Extended reading notes

Core claim

The central claim, stated in Section 10.4, is that 'a chatbot, as a conversational partner, is not an analytical thinking partner, nor can it become one with its current architecture and through text-only training.' The paper reaches this by combining four perspectives: a systems theory that views life as problem-solving, a metaphor-based account of concepts, a predictive view of perception and action, and a dual-process psychology distinguishing fast, automatic thinking from slow, analytical thinking. It introduces 'metaphorical problem propagation' as the mental space in which humans frame problems, propose solutions, and generate predictions. The paper then hypothesizes that most text use

Load-bearing premise

The load-bearing premise is that most text in LLM training datasets is 'onward text' — solution-oriented, non-reflective System 1-style prose — and that the training process encodes these as artificial metaphorical problem propagations; if training data were not dominated by such text, or if the training did not encode it in this way, the conclusion that chatbots are irredeemably System 1 would collapse.

Editorial extensions

If this is right

  • Chatbots are best understood as System 1 conversation partners: they provide plausible, fast answers that confirm a user's worldview rather than critiquing it.
  • They can still serve as creative aids — for example, finding analogies or 'dark knowledge' — but only when guided by an alert prompt writer.
  • Scaling up models or adding more text of the same kind will not produce analytical thinking; the bottleneck is the lack of experiential grounding and the reduced nature of text.
  • Safety and reliability efforts should focus on barriers, monitoring, and user vigilance, not on the expectation that chatbots will become trustworthy reasoning partners.
  • The model gives interpretability researchers a target: look for concept clusters and activation paths that correspond to problem-solution 'riverbeds.'

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A testable extension is to measure the 'onward-ness' of a training corpus and correlate it with an LLM's performance on analytical reasoning tasks; the paper's model predicts a strong negative correlation.
  • If the 'onward text' hypothesis is correct, injecting deliberately reflective, contradictory, and open-ended texts into pretraining should measurably shift chatbot behavior, even if — as the paper argues — it cannot fully overcome the structural limitation.
  • The paper's view implies that the gap between chatbot and human cognition is narrower in routine problem-solving domains (where humans also rely on System 1) and wider in open-ended analytical tasks, a differentiation the authors leave implicit.
  • The 'artificial metaphorical problem propagation' account suggests a possible evaluation metric: quantify the extent to which a chatbot's responses stay 'at the problem front' versus exploring alternative branches or revisiting assumptions.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper develops a speculative account of why basic LLM-based chatbots cannot be analytical thinking partners. It introduces 'metaphorical problem propagation' (MPP) as a synthesis of Aggregation Dynamics, Cognitive Linguistics, and Predictive Processing. It then argues that most LLM training text is 'onward text' (System 1-like, solution-oriented, non-reflective), that LLM training reconstructs 'artificial MPPs' from this text, and that consequently chatbots are confined to System-1-like thinking and cannot become analytical partners even with larger models or analytical training sets. The conclusion is an impossibility claim.

Significance. If the impossibility claim were established, it would be a significant contribution to the debate on LLM reasoning and human-AI interaction. The paper deserves credit for attempting to connect interpretability research, cognitive linguistics, and dual-process theory into a structured model, and for being explicit about its speculative character. However, the manuscript as it stands is a hypothesis sketch rather than a demonstration: the key premises are asserted, not measured, and the authors themselves disclaim the ability to assess their central model (§10.5). The paper's value lies in proposing questions and a vocabulary, not in providing a validated answer.

major comments (4)
  1. [§8.1 and §9] The premise that most training texts are 'onward text' is load-bearing but never operationalized. §8.1 lists subjective traits ('strategically oriented', 'tailored to an audience seeking comfort') with no metric; §9 simply says 'We now assume that most of the texts ... are of the previously postulated onward type.' The cited survey [50] states that narrative texts are abundant, but 'narrative' is not equivalent to 'onward' — indeed, narrative theory includes reflective, non-linear forms. Without empirical support for the dominance of onward text, the System-1 conclusion collapses.
  2. [§9 and §10.5] The inference from LLM training to 'artificial metaphorical problem propagation' is not derived. The evidence [17,48] shows that LLM activations map onto human brain responses and can represent evolving discourse situations; it does not show that these representations are 'onward' or lack reflective potential. The paper's own §10.5 admits 'we are too far removed from this highly technical field of research to assess the value of our model.' A central claim cannot rest on a model whose validity the authors disclaim.
  3. [§10.4] The claim that even an analytical training set cannot produce an analytical thinking partner depends on an unargued necessity of embodied experience for System-2 thinking. No evidence is provided that text-only training cannot acquire reflective competence; indeed, publications such as [68] indicate that prompting and training can elicit multi-step reasoning. The strong negative conclusion therefore outruns the support.
  4. [§9–§10] Circularity: §10 opens 'Suppose an LLM encodes metaphorical problem propagation, as argued above,' and then derives the System-1 conclusion from that supposition. Since the 'onward' character was built into the MPP hypothesis in §8–9, the conclusion is a restatement of the hypothesis rather than an independent result. The conclusion should be framed as a conditional, not a categorical claim.
minor comments (4)
  1. [§2.2] 'Bereska en Gavves' should be 'Bereska and Gavves' (Dutch 'en' in otherwise English text).
  2. [Reference [20]] Typo: 'Consulted om May 29, 2026' should be 'Consulted on May 29, 2026'.
  3. [§3] The term 'aggregation' is used extremely broadly ('jealousy of the gods' as a player); a formal definition or at least a clearer scope condition would help the reader follow the later argument.
  4. [Figure 3] The last line of Figure 3, 'The', appears to be a fragment; either complete the sentence or remove the stray article.

Circularity Check

1 steps flagged · score 6.0 of 10

The impossibility claim is the onward-text assumption restated under a stipulated reconstruction; §10.5 concedes the model cannot be assessed.

  1. self definitional [Section 8.1 -> Section 9 -> Section 10.2 -> Section 10.4]
    "We refer to this as onward text. ... We now assume that most of the texts in the text dataset for training publicly accessible systems such as ChatGPT, Claude and Grok are of the previously postulated onward type. ... Our hypothesis is that the training text for public chatbots has an onward character (Section 8.1). This has an effect on the metaphorical problem propagation distilled from it. It seems highly plausible that this, too, therefore has an onward character. By this we mean that a chatbot simulates a form of thinking that is more akin to System 1 thinking than to analytical thinking."

    In §8.1, 'onward text' is defined by System-1-like qualities: strategically oriented, solution-pointing, coherent, not experimental, and explicitly tied to the suspicion that much human text 'has a System 1 character.' Section 9 then assumes most LLM training text is onward. Section 10.2 concludes that the chatbot's simulated thinking is System 1 by transferring the 'onward' label to the distilled propagation ('this, too, therefore has an onward character'). The 'prediction' that a basic chatbot is System 1 is thus the assumed onwardness of the input restated under the stipulated reconstruction hypothesis; no independent mechanism or measurement is supplied. Section 10.4's denial that an analytical training set could help adds only an unargued embodied-cognition assertion, so the central i

full rationale

The paper is explicitly speculative: it labels its positive account as hypotheses, and the core inference is conditional on 'Suppose an LLM encodes metaphorical problem propagation, as argued above.' That conditionality is not itself circular. The circularity is that the central output — 'a chatbot simulates a form of thinking more akin to System 1' — is obtained by defining the training text as 'onward' (System-1-like) and then assuming the trained LLM reconstructs those propagations, so the System-1 character of the chatbot is the System-1 character of the input by construction. The chain would be a legitimate conditional argument if the premise were empirically measured and the reconstruction mechanism independently evidenced, but the paper does neither; it states the premise as an assumption and the reconstruction as a hypothesis. The self-citations [77,78] supply the AD vocabulary of finite problem types and the Innovation Illusion, but those are also framed as assumptions/conjectures rather than forced results, so I do not treat them as the primary circularity. External citations [17,48,56,80] provide some independent support for LLM representations and for the view that LLMs resemble System 1, but they do not break the definitional transfer from 'onward text' to 'System-1 chatbot.' The paper's own §10.5 admission that 'we are too far removed from this highly technical field of research to assess the value of our model' is an honest limitation, but it reinforces rather than resolves the circularity concern. Overall: one central 'prediction' reduces by construction, so the score is 6 rather than 0–2; the residual independent content is the argument that even an analytical training set would not help, which is asserted rather than derived.

Assumptions & free parameters 0 free parameters · 7 assumptions · 3 invented entities

The paper's central claims rest on conceptual assumptions from Aggregation Dynamics, Cognitive Linguistics, Predictive Processing, and dual-process psychology, plus two ad hoc hypotheses ('onward text' and 'artificial MPP') that are not empirically validated. There are no fitted numerical parameters; the free parameters are conceptual and are captured as axioms above.

assumptions (7)
  • domain assumption The 'Innovation Illusion': problem types and solution repertoires form a finite, domain-independent set of patterns.
    Assumed from Aggregation Dynamics (Section 3.3-3.4); used to argue that texts and thought are repetitive and patterned.
  • domain assumption All human thought is metaphorically structured and grounded in bodily experience (Cognitive Linguistics).
    Adopted from Lakoff and Johnson (Section 4); forms the conceptual foundation of 'metaphorical problem propagation'.
  • domain assumption Predictive Processing Theory: the brain generates predictions and perception is a 'controlled hallucination'.
    Adopted in Section 5; supplies the prediction dimension of the proposed model.
  • domain assumption Dual-process theory: System 1 is fast and automatic, System 2 is effortful and analytical.
    Used in Section 6 to classify chatbot cognition; the paper's conclusion depends on this dichotomy being a valid description of human thinking.
  • ad hoc to paper Most texts used for LLM training are 'onward text': goal-directed, coherent, non-reflective, and tailored to a comfort-seeking audience.
    Postulated in Section 8.1 and then assumed in Section 9. This is the load-bearing empirical premise for the System-1 conclusion.
  • ad hoc to paper LLM training reconstructs artificial metaphorical problem propagations from onward text.
    Stated as a hypothesis in Section 9; never independently tested, yet Section 10 treats it as the basis for the argument ('Suppose an LLM encodes...').
  • domain assumption A physical, experiencing body is required for genuine understanding; text alone cannot provide the embodied conceptual grounding needed for analytical thought.
    Adopted from embodied cognition literature in Section 10.1; drives the impossibility conclusion but is a philosophical commitment, not an empirically established fact.
invented entities (3)
  • Metaphorical problem propagation (MPP)
    purpose: A model of human thought space combining problem positions, conceptual metaphors, and predictive processing; intended to explain human flexibility and chatbot limitations.
    Introduced in Section 7. It is a conceptual construct with no direct measurement or falsifiable handle. The cited neuroimaging and representation studies ([17,48]) are offered as 'indications', not as tests of MPP itself.
  • Artificial metaphorical problem propagation
    purpose: The internal structure hypothesized to be reconstructed in an LLM during training and activated during conversation.
    Section 9 posits this as the mechanism behind chatbot behaviour, but gives no operational definition or method to detect it. As stated, it cannot be confirmed or refuted.
  • Onward text
    purpose: A category of text claimed to dominate LLM training corpora and to encode reduced, System-1-like thinking.
    Defined in Section 8.1. No dataset statistics, annotation scheme, or quantitative criterion are provided; the prevalence of 'onward text' is asserted rather than demonstrated.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Some hypotheses on how chatbots work in problem-solving-driven conversations. Large Language Models as confirmation of the Innovation Illusion." pith.science (2026). https://pith.science/paper/ZZX3ATEB

@misc{pith2026260607722,
  author       = {Pith},
  title        = {Pith review of: Some hypotheses on how chatbots work in problem-solving-driven conversations. Large Language Models as confirmation of the Innovation Illusion},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZZX3ATEB}},
  note         = {Machine review of arXiv:2606.07722}
}
read the original abstract

We discuss the nature of chatbots as conversation partners in problem-solving. What can chatbots do and what can't they do? We develop hypotheses on how this can this be explained. Our argument draws on insights from Aggregation Dynamics, Cognitive Linguistics, Neuropsychology and Psychology. We establish that chatbots are multifaceted and composite systems. Our argument focuses on basic chatbots in the hope of thereby making statements about the core functionality of more advanced chatbots. Basic chatbots are assumed to consist of a Large Language Model (LLM) with a simple interface. The main results of our research are: a description of human imagination, understanding and thinking based on so-called metaphorical problem propagations; the hypothesis that the texts in the text dataset used for training LLMs have specific characteristics and that these texts only partially imitate human thinking and understanding; the hypothesis that the LLM training process encodes artificial metaphorical problem propagations into an LLM from these text datasets. Our conclusions are that a basic chatbot cannot be a thinking partner capable of matching the cognitive flexibility of humans, and that further development of the Large Language Model will not lead to this either. But chatbots exist, they are being used on a massive scale, by both individuals and organisations. It is therefore socially and politically important to understand them. Our article aims to contribute to the discussion on the functioning, benefits and drawbacks of chatbots. Cognitive Linguistics shows how the use of metaphor is an expression of our thinking. Aggregation Dynamics, is an attempt at a comprehensive systems theory. We believe that the concept of metaphorical problem propagation could provide an interesting addition for both. Chatbots a solution? For what?

Figures

Figures reproduced from arXiv: 2606.07722 by the authors.

Figure 1
Figure 1. An example of abstract problem propagation. Here, the p’s represent [PITH_FULL_IMAGE:figures/full_fig_p011_1.png] view at source ↗
Figure 1
Figure 1. An example of abstract problem propagation. Here, the p’s represent [PITH_FULL_IMAGE:figures/full_fig_p010_1.png] view at source ↗
Figure 2
Figure 2. An example of problem propagation and a suggestion of some context [PITH_FULL_IMAGE:figures/full_fig_p013_2.png] view at source ↗
Figures from the paper (3 more)
Figure 2
Figure 2. Figure 2: An example of problem propagation and a suggestion of some context [PITH_FULL_IMAGE:figures/full_fig_p012_2.png]
Figure 3
Figure 3. Figure 3: A linearisation of the problem propagation shown in Figure 2, in [PITH_FULL_IMAGE:figures/full_fig_p024_3.png]
Figure 3
Figure 3. Figure 3: A linearisation of the problem propagation shown in Figure 2, in [PITH_FULL_IMAGE:figures/full_fig_p023_3.png]

Review history (4 revisions) →

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

79 extracted references

  1. [56]

    Philipp Mondorf and Barbara Plank,Beyond Accuracy: Evaluating the Reasoning Behavior of Large Language Models – A Survey, 2024, arXiv, DOI

    2024

  2. [80]

    Webster,Natural and artificial intelligence – the psychotechnical agenda of the 21st century, Journal of Psychology and AI1(2025), no

    Craig S. Webster,Natural and artificial intelligence – the psychotechnical agenda of the 21st century, Journal of Psychology and AI1(2025), no. 1, 2491445, DOI

    2025

  3. [50]

    Liu, Aditya Joshi, and Paul Dawson,Narrative Theory-Driven LLM Methods for Automatic Story Generation and Understanding: A Sur- vey, 2026, arXiv, DOI

    David Y. Liu, Aditya Joshi, and Paul Dawson,Narrative Theory-Driven LLM Methods for Automatic Story Generation and Understanding: A Sur- vey, 2026, arXiv, DOI. 38

    2026

  4. [17]

    Charlotte Caucheteux, Alexandre Gramfort, and Jean-Remi King,Evi- dence of a predictive coding hierarchy in the human brain listening to speech, Nature Human Behaviour7(2023), no. 3, pp. 430–441, DOI

    2023

  5. [48]

    Belinda Z. Li, Maxwell Nye, and Jacob Andreas,Implicit representations of meaning in neural language models, Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers) (Online) (Chengqing Zong, Fei Xia, Wenjie Li, and Roberto Nav...

    2021

  6. [68]

    Surv.58(2025), no

    Aske Plaat, Annie Wong, Suzan Verberne, Joost Broekens, Niki Van Stein, and Thomas B¨ ack,Multi-Step Reasoning with Large Language Models, a Survey, ACM Comput. Surv.58(2025), no. 6, DOI

    2025

  7. [5]

    Het begin van het computertijdperk in Nederland, AUP, Amsterdam, 2016, In Dutch

    Gerard Alberts and Bas van Vlijmen,Computerpioniers. Het begin van het computertijdperk in Nederland, AUP, Amsterdam, 2016, In Dutch

    2016

  8. [6]

    Tomas Humberto Montiel Alcantara, David Kr¨ utli, Revathi Ravada, and Thomas Hanne,Multilingual Text Summarization for German Texts Using Transformer Models, Information14(2023), no. 6

    2023

Show all 79 references
  1. [7]

    Bailey,When push comes to shove: A computational model of the role of motor control in the acquisition of action verbs, Ph.D

    David R. Bailey,When push comes to shove: A computational model of the role of motor control in the acquisition of action verbs, Ph.D. thesis, EECS Department, University of California, Berkeley, 1997

    1997

  2. [8]

    Lochan Basyal and Mihir Sanghvi,Text summarization using large language models: A comparative study of mpt-7b-instruct, falcon-7b-instruct, and openai chat-gpt models, ArXiv (2023), DOI

    2023

  3. [9]

    Townsend,Ecology: From individuals to ecosystems, 5e ed., Wiley Blackwell, 2021

    Michael Begon and Colin R. Townsend,Ecology: From individuals to ecosystems, 5e ed., Wiley Blackwell, 2021

    2021

  4. [10]

    Emily M. Bender, Timnit Gebru, Angelina McMillan-Major, and Shmar garet Shmitchell,On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?, Conference on Fairness, Accountability, and Trans- parency, ACM F AccT ’21, 2021, DOI. 35

    2021

  5. [11]

    Leonard Bereska and Stratis Gavves,Mechanistic Interpretability for AI Safety – A Review, Transactions on Machine Learning Research (2024), URL

    2024

  6. [12]

    Eric Bigelow Bigelow and Tomer Ullman,People Evaluate Agents Based on the Algorithms That Drive Their Behavior, Open Mind9(2025), 1411– 1430, DOI

    2025

  7. [13]

    1 & 2, Clarendon, Oxford University Press, 2006

    Margaret Boden,Mind As Machine: A History Of Cognitive Science, vol. 1 & 2, Clarendon, Oxford University Press, 2006

    2006

  8. [14]

    Margaret Boden,Artificial Intelligence, Oxford University Press, 2018

    2018

  9. [15]

    John Brockman (ed.),What to Think about Machines that Think, Harper Perennial, 2015

    2015

  10. [16]

    Alexander Cameron,What is 4E cognitive science?, Phenomenology and the Cognitive Sciences (2025), DOI

    2025

  11. [18]

    How Our Minds Predict and Shape Reality, Penguin, 2024 (2023)

    Andy Clark,The Experience Machine. How Our Minds Predict and Shape Reality, Penguin, 2024 (2023)

    2024

  12. [19]

    Dewey,Don ’t Breathe the Air: Air Pollution and U.S

    S.H. Dewey,Don ’t Breathe the Air: Air Pollution and U.S. Environmental Politics, 1945-1970, Environmental History Series, no. 16, Texas A&M University Press, 2000

    1945

  13. [20]

    Consulted om May 29, 2026

    de.wikipedia.org,Paul Watzlawick, URL. Consulted om May 29, 2026

    2026

  14. [21]

    Aniket Didolkar, Anirudh Goyal, Nan Rosemary Ke, Siyuan Guo, Michal Valko, Timothy Lillicrap, Danilo Rezende, Yoshua Bengio, Michael Mozer, and Sanjeev Arora,Metacognitive capabilities of LLMs: an exploration in mathematical problem solving, Proceedings of the 38th Internation...

    2024

  15. [22]

    September 2025, pp

    Maximilian Dreyer, Jim Berend, Tobias Labarta, Johanna Vielhaben, Thomas Wiegand, Sebastian Lapuschkin, and Wojciech Samek,Mecha- nistic understanding and validation of large AI models with SemanticLens, Nature Machine Intelligence7, no. September 2025, pp. 1572–1585, DOI

    2025

  16. [23]

    In Dutch, translated from French by Hans E

    Giuliano da Empoli,Het uur van de wolven, Atlas Contact, 2025, Original title:L’Heure des pr´ edateurs. In Dutch, translated from French by Hans E. van Riemsdijk

    2025

  17. [24]

    Conceptual blending and the mind’s hidden complexities, Basic Books, 2002

    Gilles Fauconnier and Mark Turner,The way we think. Conceptual blending and the mind’s hidden complexities, Basic Books, 2002. 36

    2002

  18. [25]

    Feldman,From Molecule to Metaphor

    Jerome A. Feldman,From Molecule to Metaphor. A Neural Theory of Lan- guage, MIT Press, 2006

    2006

  19. [26]

    Elliot Glazer, Ege Erdil, Tamay Besiroglu, Diego Chicharro, Evan Chen, Alex Gunning, Caroline Falkman Olsson, Jean-Stanislas Denain, Anson Ho, Emily de Oliveira Santos, Olli J¨ arviniemi, Matthew Barnett, Robert Sandler, Matej Vrzala, Jaime Sevilla, Qiuyu Ren, Elizabeth Pratt,...

    2025

  20. [27]

    Hugo Gon¸ calo Oliveira, Brad Spendlove, Pablo Gerv´ as, and Dan Ventura (eds.),Proceedings of the Sixteenth International Conference on Computa- tional Creativity, Association for Computational Creativity, 2025, PDF

    2025

  21. [28]

    Kazjon Grace, Maria Teresa Llano, Pedro Martins, and Maria Hedblom (eds.),Proceedings of the Fifteenth International Conference on Computa- tional Creativity, Association for Computational Creativity, 2024, PDF

    2024

  22. [29]

    Heyes,New thinking: the evolution of human cognition, Philosophical Transactions of the Royal Society (2012), pp

    C. Heyes,New thinking: the evolution of human cognition, Philosophical Transactions of the Royal Society (2012), pp. 2091–6, August 5, 367(1599), Biological Science, DOI

    2012

  23. [30]

    Analogy as the Fuel and Fire of Thinking, Basic Books, 2013

    Douglas Hofstadter and Emmanuel Sander,Surfaces and Essences. Analogy as the Fuel and Fire of Thinking, Basic Books, 2013

    2013

  24. [31]

    Holyoak, Nicholas Ichien, and Hongjing Lu,Analogy and the Gen- eration of Ideas, Creativity Research Journal36(2024), no

    Keith J. Holyoak, Nicholas Ichien, and Hongjing Lu,Analogy and the Gen- eration of Ideas, Creativity Research Journal36(2024), no. 3, 532–543, DOI

    2024

  25. [32]

    Kelly,Smogtown: The Lung-Burning History of Pollution in Los Angeles, Overlook Press, Woodstock & New York, 2008

    Chip Jacobs and William J. Kelly,Smogtown: The Lung-Burning History of Pollution in Los Angeles, Overlook Press, Woodstock & New York, 2008

    2008

  26. [33]

    1, 155–205, DOI

    Di Jin, Zhijing Jin, Zhiting Hu, Olga Vechtomova, and Rada Mihalcea, Deep learning for text style transfer: A survey, Computational Linguistics 48(2022), no. 1, 155–205, DOI

    2022

  27. [34]

    How our bodies give rise to understanding, The University of Chicago Press, 2017

    Mark Johnson,Embodied mind, meaning and reason. How our bodies give rise to understanding, The University of Chicago Press, 2017

    2017

  28. [35]

    James Joyce,Finnegans Wake, Wordsworth Editions, 2012 (1939)

    2012

  29. [36]

    Daniel Kahneman,Thinking, Fast and Slow, Farrar, Straus & Giroux Inc, 2012

    2012

  30. [37]

    The New Science of Animal Com- munication, Penguin, 2025

    Arik Kershenbaum,Why Animals Talk. The New Science of Animal Com- munication, Penguin, 2025. 37

    2025

  31. [38]

    Jon Kleinberg and Sendhil Mullainathan,We built them, but we don ’t un- derstand them, What to Think about Machines that Think (John Brock- man, ed.), Harper Perennial, 2015

    2015

  32. [39]

    Koyejo, S

    Takeshi Kojima, Shixiang (Shane) Gu, Machel Reid, Yutaka Matsuo, and Yusuke Iwasawa,Large language models are zero-shot reasoners, Ad- vances in Neural Information Processing Systems (S. Koyejo, S. Mohamed, A. Agarwal, D. Belgrave, K. Cho, and A. Oh, eds.), vol. 35, Curran Ass...

    2022

  33. [40]

    September 20-21, Achtergrond pp

    Marloes de Koning,ChatGPT leek vol begrip over Adams zelfmoordplan, NRC (2025), no. September 20-21, Achtergrond pp. 4–6, in Dutch

    2025

  34. [41]

    Kumar,Large language models (LLMs): survey, technical frameworks, and future challenges, Artificial Intelligence Review57(2024), p

    P. Kumar,Large language models (LLMs): survey, technical frameworks, and future challenges, Artificial Intelligence Review57(2024), p. 260 et seq., DOI

    2024

  35. [42]

    When we merge with AI, The Bodley Head, Penguin, 2024

    Ray Kurzweil,The singularity is nearer. When we merge with AI, The Bodley Head, Penguin, 2024

    2024

  36. [43]

    George Lakoff,The Neural Theory of Metaphor, 2009 (2008), UC Berkeley Previously Published Works, URL

    2009

  37. [44]

    ed., The University of Chicago Press, Chicago, IL, 2003 (1980)

    George Lakoff and Mark Johnson,Metaphors we live by, [with a new after- word]. ed., The University of Chicago Press, Chicago, IL, 2003 (1980)

    2003

  38. [45]

    Yann LeCun and Courant,A path towards autonomous machine intelli- gence version 0.9.2, 2022-06-27, 2022, URL

    2022

  39. [46]

    Taewhoo Lee, Minju Song, Chanwoong Yoon, Jungwoo Park, and Jaewoo Kang,The curious case of analogies: Investigating analogical reasoning in large language models, 2025, arXiv, DOI

    2025

  40. [47]

    Martha Lewis and Melanie Mitchell,Evaluating the robustness of analogi- cal reasoning in large language models, Transactions on Machine Learning Research (2025), OpenReview.net, URL

    2025

  41. [49]

    Litseller.com,Finnegans Wake, consulted on April 13, 2026, URL

    2026

  42. [51]

    Yang Liu, Jiahuan Cao, Chongyu Liu, Kai Ding, and Lianwen Jin,Datasets for large language models: a comprehensive survey, Artificial Intelligence Review58(2025), article number 403, DOI

    2025

  43. [52]

    Op zoek naar het bewustzijn van mens, dier en plant (en misschien zelfs AI), Uitgeverij Balans, 2025, in Dutch

    Sebastiaan Mathˆ ot,Een wereld vol denkers. Op zoek naar het bewustzijn van mens, dier en plant (en misschien zelfs AI), Uitgeverij Balans, 2025, in Dutch

    2025

  44. [53]

    Maturana and Francisco J

    Humberto R. Maturana and Francisco J. Varela,The tree of knowledge: The biological roots of human understanding, revised ed., Shambhala Pub- lications Inc., Boston (Mass.), 1998

    1998

  45. [54]

    65–93, Springer International Publishing, 2018, European Studies in Philosophy of Science,volume 9, DOI

    Anne Sophie Meincke,Bio-Agency and the Possibility of Artificial Agents, pp. 65–93, Springer International Publishing, 2018, European Studies in Philosophy of Science,volume 9, DOI

    2018

  46. [55]

    A Guide for Thinking Humans, paperback, 2020 ed., Pelican Books & Penguin Random House, 2019

    Melanie Mitchell,Artificial Intelligence. A Guide for Thinking Humans, paperback, 2020 ed., Pelican Books & Penguin Random House, 2019

    2020

  47. [57]

    Muraki and Penny M

    Emiko J. Muraki and Penny M. Pexman,The role of emotion in acquisition of verb meaning, Cognition and Emotion39(2025), no. 7, 1457–1464, DOI

    2025

  48. [58]

    16366–16393, URL

    Tarek Naous, Michael J Ryan, Alan Ritter, and Wei Xu,Having Beer after Prayer? Measuring Cultural Bias in Large Language Models, Proceedings of the 62nd Annual Meeting of the Association for Computational Linguis- tics (Volume 1: Long Papers) (Bangkok, Thailand) (Lun-Wei Ku, A...

    2024

  49. [59]

    Arvind Narayanan and Sayash Kapoor,AI Snake Oil: what artificial in- telligence can do, what it can ’t, and how to tell the difference, Princeton University Press, 2024

    2024

  50. [60]

    Simon, Echo Point Books & Media, 2019 (1972)

    Allen Newell and Herbert A. Simon, Echo Point Books & Media, 2019 (1972)

    2019

  51. [61]

    2, 349–366, DOI

    Doroteya Nikolova,Predictive Processing Theory in Mind Studies: Cross Points with 4e Cognition and Cognitive Linguistics, Open Journal of Phi- losophy15(2025), no. 2, 349–366, DOI

    2025

  52. [62]

    Stevenson,Analogical reasoning inside large language models: Concept vectors and the limits of abstraction, 2025, arXiv, DOI

    Gustaw Opie lka, Hannes Rosenbusch, and Claire E. Stevenson,Analogical reasoning inside large language models: Concept vectors and the limits of abstraction, 2025, arXiv, DOI

    2025

  53. [63]

    Andrew Ortony (ed.),Metaphor and Thought, second ed., Cambridge Uni- versity Press, 1993 (1979). 39

    1993

  54. [64]

    1263– 1280, DOI

    Matteo Pasquinelli and Vladan Joler,The Nooscope manifested: AI as instrument of knowledge extractivism, AI & SOCIETY36(2021), pp. 1263– 1280, DOI

    2021

  55. [65]

    Jonathon Phillips, Carina A

    P. Jonathon Phillips, Carina A. Hahn, Peter C. Fontana, Amy N. Yates, Kristen Greene, David A. Broniatowski, and Mark A. i Przybock,Four Principles of Explainable Artificial Intelligence, Tech. Report Interagency or Internal Report 8312, National Institute of Standards and Tec...

    2021

  56. [66]

    Streven naar veiligheid in een wereld vol risico en onzekerheid, Boom juridische uitgevers, 2008, in Dutch

    Roel Pieterman,De voorzorgcultuur. Streven naar veiligheid in een wereld vol risico en onzekerheid, Boom juridische uitgevers, 2008, in Dutch

    2008

  57. [67]

    298–311, URL

    Andrew Piper, Richard Jean So, and David Bamman,Narrative Theory for Computational Narrative Understanding, Proceedings of the 2021 Con- ference on Empirical Methods in Natural Language Processing (Online and Punta Cana, Dominican Republic) (Marie-Francine Moens, Xuanjing Huan...

    2021

  58. [69]

    Karl Popper,All Life is Problem Solving, Routlegde, 2001 (1999)

    2001

  59. [70]

    23993–24010, URL

    Chengwei Qin, Wenhan Xia, Tan Wang, Fangkai Jiao, Yuchen Hu, Bosheng Ding, Ruirui Chen, and Shafiq Joty,Relevant or Random: Can LLMs Truly Perform Analogical Reasoning?, Findings of the Association for Compu- tational Linguistics: ACL 2025 (Vienna, Austria) (Wanxiang Che, Joyc...

    2025

  60. [71]

    Herman de Regt and Hans Dooremalen,Het snapgevoel: Hoe de illusie van begrip ons denken gijzelt, Boom, Amsterdam, 2015, in Dutch

    2015

  61. [72]

    A new science of consciousness, paperback 2022 ed., Faber & Faber ltd., 2021

    Anil Seth,Being you. A new science of consciousness, paperback 2022 ed., Faber & Faber ltd., 2021

    2022

  62. [73]

    Reddy, Neural Abstractive Text Summarization with Sequence-to-Sequence Models, ACM/IMS Trans

    Tian Shi, Yaser Keneshloo, Naren Ramakrishnan, and Chandan K. Reddy, Neural Abstractive Text Summarization with Sequence-to-Sequence Models, ACM/IMS Trans. Data Sci.2(2021), no. 1, DOI

    2021

  63. [74]

    Consulted on April 13, 2026

    Tensorflow.org,Tensorflow C4 dataset, URL. Consulted on April 13, 2026

    2026

  64. [75]

    Transmathematica, Consulted on May 21, 2026, see: URL

    2026

  65. [76]

    Jan Verplaetse,Zonder vrije wil: Een filosofisch essay over verantwoordeli- jkheid, Uitgeverij Nieuwezijds, 2011. 40

    2011

  66. [77]

    van Vlijmen,Exploration of Problem and Solution Types in Los Angeles County Smog Control 1943-1976, Transmathematica (2025), DOI

    S.F.M. van Vlijmen,Exploration of Problem and Solution Types in Los Angeles County Smog Control 1943-1976, Transmathematica (2025), DOI

    1943

  67. [78]

    van Vlijmen,Problem Positions and Nineteen Instances in the Smog Control Case Los Angeles County 1943-1976, Transmathematica (2025), DOI

    S.F.M. van Vlijmen,Problem Positions and Nineteen Instances in the Smog Control Case Los Angeles County 1943-1976, Transmathematica (2025), DOI

    1943

  68. [79]

    7097–7135, URL

    Mengru Wang, Yunzhi Yao, Ziwen Xu, Shuofei Qiao, Shumin Deng, Peng Wang, Xiang Chen, Jia-Chen Gu, Yong Jiang, Pengjun Xie, Fei Huang, Huajun Chen, and Ningyu Zhang,Knowledge Mechanisms in Large Lan- guage Models: A Survey and Perspective, Findings of the Association for Comput...

    2024

  69. [81]

    Consulted on May 21, 2026

    wikipedia.org,Houston, we have a problem, URL. Consulted on May 21, 2026

    2026

  70. [82]

    and Why Does It Work?, Kindle ed., Wolfram Media, Inc., 2023

    Stephen Wolfram,What Is ChatGPT Doing ... and Why Does It Work?, Kindle ed., Wolfram Media, Inc., 2023

    2023

  71. [83]

    656–668, DOI

    Xiao Yu, Zexian Zhang, Feifei Niu, Xing Hu, Xin Xia, and john Grundy, What Makes a High-Quality Training Dataset for Large Language Mod- els: A Practitioners’ Perspective, Proceedings of the 39th IEEE/ACM International Conference on Automated Software Engineering, ASE ’24, IEE...

    2024

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.