Submitted:
09 December 2025
Posted:
14 December 2025
You are already at the latest version
Abstract
Keywords:
Part I: Introduction—A Decisive Anomaly: Why Interfaces Rather Than Stories Become Fixed in Dreams?
0.1. The Posing of the Problem: The Technological Unconscious Turn in Cognitive Science
0.2. The Hegemony and Fundamental Dilemma of the Process-Centered Paradigm
0.3. This Paper’s Argumentative Path and Core Thesis: Towards a Spatiotemporal Dynamics of Cognition
- Core mechanism is Interaction Architecture Internalization: In goal-oriented repetitive interactions, the cognitive system strips away specific content, extracts the abstract logic and temporal structure of the interaction (i.e., the "architecture"), and solidifies it into internal basic cognitive structures.
- Core variable is Learning Time Delay Dose: Architecture internalization does not necessarily occur; it is driven by a quantifiable physical variable—namely, the time interval between an individual initiating an "exploratory operation" aimed at achieving cognitive closure and receiving "deterministic feedback" of adjudicative significance, and the "dose" D accumulated through effective repetitions. When the dose D exceeds an individual-specific critical threshold , the cognitive system undergoes a phase transition, and the external architecture becomes irreversibly internalized.
- Unified explanatory power: This model, with "time delay" as the order parameter and "architecture internalization" as the core mechanism, will demonstrate powerful unified explanatory power. It will not only satisfactorily explain the dream anomaly in the Mayer report but also provide a coherent, mechanistic explanatory framework for a range of phenomena spanning from language acquisition to meme propagation, and from personality formation to technological addiction.
1. Integration of Theoretical Foundations: From Piaget and Chomsky to Einstein and Landau
1.1. The Legacy and Limitations of Piaget: The Dynamic Refinement of Internalization
1.2. Chomsky’s Decisive Critique: From Linguistic Architecture to Interaction Architecture
1.3. Einstein’s Philosophical Enlightenment: The Spatiotemporal Revolution in Cognitive Science
- Relativity of Simultaneity: Einstein demolished the Newtonian notion of an absolute, universal flow of time. He showed us that the measurement of time intervals depends on the observer’s reference frame and state of motion. This principle compels us to abandon treating "learning time delay" as an absolute, background parameter external to the cognitive system. Instead, we must elevate it to a core, relativistic state variable within the cognitive system. The perception and effect of time delay depend on the system’s current state (e.g., motivation level, expectation, attention); it is itself part of the cognitive dynamics, not an externally given physical quantity.
- Equivalence Principle: This is the most direct physical analogy and philosophical basis for our Learning Time Delay Equivalence Principle. Einstein demonstrated that inside a closed elevator, no experiment can distinguish between the effects of a gravitational field and those of accelerated motion. Similarly, we assert: In the phase transition process driving interaction architecture internalization, no experiment confined within the system can distinguish whether the fundamental driving factor is the "objective learning time delay dose (D)" itself, or any microscopic cognitive activity (such as implicit reasoning, memory retrieval, emotional fluctuation) accompanying it within that time window. It is the objective fact of "waiting"—not the subjective content of "what was thought" during the wait—that constitutes the equivalent cause of architecture internalization. This principle has profound methodological significance: it forces us to recognize that subjective cognitive states can be equated with an objectively measurable physical quantity—the time delay dose—thus introducing unprecedented objectivity and computability into cognitive science.
1.4. Wiener’s Cybernetics and Landau’s Mathematics: From Feedback Timing to Phase Transition Order
- Wiener’s Cybernetics and Feedback Timing: Cybernetics unifies organisms and machines as systems regulated through information feedback. However, contemporary cognitive science, in applying this framework, has focused almost exclusively on the informational content of feedback, systematically ignoring its timing. Our model redefines learning time delay—this decisive yet long-neglected variable in the feedback loop—as the core element of cognitive dynamics. It is not what the feedback "says," but when it arrives, that plays the more fundamental role in shaping cognitive structure.
- Landau’s Phase Transition and Hale’s Bifurcation Mathematics: To understand how cognitive structures "suddenly" emerge from seemingly disordered interactions, we need the powerful metaphor and mathematical tools of Landau’s phase transition theory. Landau showed us that when an order parameter (in our model, the learning time delay dose D) crosses a critical threshold, the system undergoes a phase transition, microscopic symmetry is broken, and a new macroscopic order is born. This provides an appropriate, not merely poetic, mathematical framework for understanding the abrupt formation of cognitive structures. Furthermore, the bifurcation theory of delay differential equations, established by J.K. Hale and others, mathematically rigorously proves that time delay itself can serve as a bifurcation parameter for a system; when the delay exceeds a certain critical value, the system loses stability and enters a completely new dynamical regime. This provides indisputable mathematical support for the hypothesis of "learning time delay as the order parameter for cognitive phase transitions," elevating our model from a phenomenological description to the level of a computable dynamical system.
2. Core Concepts and First Principles: The Formalization of Architecture, Time Delay, and Internalization
2.1. Operational Definitions of Core Concepts
-
1. Interaction Architecture
- -
- Definition: The stable, dynamic formal structure governing the progression of interaction within a goal-oriented context. It explicitly specifies a set of permissible operation sequences, the rule sets these operations follow, and the temporal relationship between actions and feedback.
- -
- Essence: It is the “grammar” of interaction, a system of invariant relations independent of any specific informational payload.
- -
-
Examples:
- *
- The interaction architecture with ChatGPT is: “User inputs a textual query → System displays a ’Thinking...’ state (time delay ) → System generates a response word-by-word as a text stream.”
- *
- The interaction architecture of a phone call is: “Caller dials number → System establishes connection (time delay) → Bidirectional, real-time, audio-only conversation stream.”
- *
- The architecture for a rat pressing a lever in a Skinner box is: “Press lever (operation) → Food pellet drops (deterministic feedback).”
-
2. Interaction Content
- -
- Definition: The specific informational filler that is produced, transmitted, or processed during a single instantiation of a particular interaction architecture.
- -
- Essence: It is the “semantics” of interaction, the probabilistic, context-dependent specific product.
- -
-
Examples:
- *
- The specific question posed to ChatGPT (“Explain relativity”) and the generated answer text.
- *
- The specific topic discussed during a phone call.
- *
- The internal state of hunger or satiety of the rat when pressing the lever (this belongs to its “interaction” content, not the architecture).
-
3. Learning Time Delay ()
- -
- Definition: In goal-oriented interaction, the time interval between an individual initiating an exploratory operation aimed at achieving cognitive closure or a goal, and the onset of the deterministic feedback signal, generated by that operation, which holds adjudicative significance. Denoted as .
- -
-
Operational Delineation:
- *
- Start Point: The execution of a complete request or action (e.g., pressing Enter in the ChatGPT dialog box).
- *
- End Point: The initial appearance of the deterministic feedback signal (e.g., the first word of ChatGPT’s answer beginning to generate, not the complete delivery of the answer).
- -
- Effective Delay Threshold (): We postulate a psychophysical effective threshold; delays below this threshold (e.g., delays in normal human conversation) contribute negligibly to architecture internalization. Significant internalization effects occur when .
-
4. Learning Time Delay Dose (D)
- -
- Definition: The cumulative time-delay exposure that drives the cognitive system’s phase transition of architecture internalization. It is a function of the single effective learning time delay and the number of effective repetitions n, denoted as .
- -
- Preliminary Model: As a first approximation, we can assume a linear weighting model: , where the sum is over all effective interactions, and is the weight for each interaction (potentially modulated by factors like attention, motivation). The contribution is zero if .
-
5. Cognitive Phase Transition and Critical Threshold ()
- -
- Definition: A qualitative, abrupt leap of the cognitive system from one cognitive state to another. Here, it specifically refers to the transition from the state of not having internalized a particular interaction architecture to the state where that interaction architecture has become a dominant internal structure.
- -
- Critical Threshold (): The critical value of the learning time delay dose D required to trigger the internalization phase transition for a specific interaction architecture. This threshold may vary based on individual, architectural complexity, emotional state, and other factors.
2.2. Formal Statement of the First Principles
-
Principle 1: Interaction Architecture Internalization Principle
- -
- Statement: In goal-oriented repetitive interaction, when the effective learning time delay dose D accumulates to the critical value for that specific architecture in that individual, the cognitive system necessarily undergoes a phase transition. The hallmark of this transition is that the external interaction architecture is extracted from the specific content it carries and is irreversibly solidified into a stable internal structure that dominates the form of subsequent cognitive outputs (including thought, imagination, and even dreams).
- -
- Implication: This principle establishes the cognitive priority of “form over content.” It predicts that once the phase transition occurs, an individual’s cognitive activity will spontaneously follow the logic of the internalized architecture, just as a language user unconsciously follows internalized grammar.
-
Principle 2: Learning Time Delay Equivalence Principle
- -
- Statement: In the phase transition process driving interaction architecture internalization, any attempt to distinguish, through experiments or introspection confined to the system itself, whether the fundamental driving factor is the “objective learning time delay dose D” itself, or any “micro-cognitive activity” (such as implicit reasoning, imagination, memory association, or emotional fluctuation) accompanying it within that time window, is in principle impossible.
- -
- Implication: This principle is a direct mapping of Einstein’s Equivalence Principle into cognitive science. It means that for the effect of architecture internalization, D is the sole equivalent order parameter. We cannot, and need not, inquire into what the brain specifically did during the “three seconds of waiting”; the objective delay of these three seconds itself is the necessary and sufficient condition driving the phase transition. This greatly simplifies the complexity of causality and anchors the root cause of cognitive structural change to an objectively measurable physical quantity.
2.3. The Learning Time Delay Dose Driven Model: A Formal Framework
3. Systematic Critique I: The Poverty of Dream Theories and the Absence of Architectural Internalization
3.1. Critique of the Activation-Synthesis Hypothesis: The Pallor of Randomness and the Absence of Structure
- Core Tenet Recap: Proposed by J. Allan Hobson and Robert McCarley, this hypothesis posits that the essence of dreaming originates from random neural signals periodically emitted by the brainstem (particularly the pons) during REM sleep. This bottom-up "activation" bombards the cerebral cortex, and higher cognitive centers are forced to "synthesize" these chaotic signals into barely coherent narratives and imagery. Dreams are thus seen as a "cognitive hallucination," a byproduct of the brain seeking meaning in noise.
- Attempted Explanation and Failure regarding the Mayer Phenomenon: According to this hypothesis, the recurrent appearance of the ChatGPT interface in dreams should be interpreted as the brainstem’s random activation accidentally triggering cortical areas related to visual interfaces and text input, with the synthesis mechanism cobbling these elements into a vague scene about using AI.
-
Analysis of Fundamental Flaws:
- The Specificity Paradox: The core of this hypothesis lies in "randomness." Yet, the data from the Mayer report shows astonishing specificity and consistency—up to 93% of AI-related nightmares fixate on the same interaction interface. How could random, highly variable brainstem activation systematically and across individuals synthesize the exact same, highly specific interaction architecture? This is akin to expecting a hurricane sweeping through a junkyard to repeatedly assemble the same precise computer. The probability is logically low enough to refute random activation as the primary driver.
- Neglect of Architectural Precision: The hypothesis completely fails to explain why the synthesized element is the formal framework of the interaction (the dialog box, the waiting state) rather than more emotionally charged content (e.g., AI rebellion or comfort). It only explains where the "materials" of the dream might come from but is utterly powerless to explain why these materials adhere to such a precise structural logic. Random activation can produce image fragments, but it cannot produce the grammar governing the sequence of images.
3.2. Critique of the Content Reorganization Hypothesis: The Misjudgment of "Significant" Content
- Core Tenet Recap: This hypothesis locates the function of dreaming in the offline processing of memory, specifically the integration, consolidation, and assimilation of significant daily experiences (especially emotional, unresolved conflicts) into existing long-term memory networks. Dreaming is the brain’s "nocturnal workshop," processing the informational "raw materials" ingested during wakefulness.
- Attempted Explanation and Failure regarding the Mayer Phenomenon: Proponents of this hypothesis might argue that interaction with ChatGPT has become a "significant" daily experience for modern humans and thus becomes material for dream processing.
-
Analysis of Fundamental Flaws:
- Confusion of Content and Architecture: This hypothesis commits a fundamental category mistake. It presupposes that what is reorganized is the content of experience. However, Mayer’s data clearly indicates that what is preferentially "reorganized" is not the semantic content of the interaction (what was discussed) but its formal architecture (how the interaction occurred). What the brain diligently consolidates at night is not the specific topics of conversation with AI, but the "question-wait-answer" interaction protocol itself. The hypothesis suffers from a fatal ambiguity on the fundamental question of "what constitutes significant content," failing to explain why the form of interaction would be prioritized for processing during precious offline resources over the substance of the interaction.
- Deviation from the Emotional Core: The hypothesis often emphasizes the processing of emotional memories. Yet, in ChatGPT interactions, the strongest emotions likely arise from the dialogue content itself (e.g., receiving a brilliant idea or a terrifying answer). But the dream "chooses" the relatively emotionally neutral interface as its core. This strongly suggests that the mechanism driving internalization is independent of, and perhaps even prior to, traditional emotional salience.
3.3. Critique of the Continuity Hypothesis: Circularity and the Selectivity Problem
- Core Tenet Recap: This is an intuitive and popular hypothesis proposing that dreams exhibit continuity in themes, concerns, and content with an individual’s waking thoughts and behaviors. The so-called "what one thinks about by day, one dreams about by night."
- Attempted Explanation and Failure regarding the Mayer Phenomenon: This hypothesis seems to offer the most direct explanation: people use ChatGPT extensively during the day, so they dream about it at night.
-
Analysis of Fundamental Flaws:
- The Trap of Circular Reasoning: The Continuity Hypothesis immediately falls into circular reasoning when explaining architectural internalization. It claims the ChatGPT interface appears in dreams because one thinks/uses it while awake. But this merely pushes the question back one step: Why does waking "thought" about ChatGPT so 集中地, selectively manifest as a concern with the form of its interaction interface, rather than with its capabilities, social impact, or ethical implications? The hypothesis uses the empty label "continuity" to obscure the real question needing explanation—namely, in which dimension does continuity specifically manifest?
- Failure of Feature Selectivity: The hypothesis lacks any principled mechanism to determine which of the infinite features of waking experience get "continued" into dreams. When interacting with ChatGPT, we simultaneously experience its content, its interface, its speed, its utility, its social meaning... Why is it only that interaction architecture that is so faithfully "continued"? The Continuity Hypothesis itself cannot answer this selectivity problem; it can only describe post hoc, not predict.
3.4. Conclusion
4. Systematic Critique II: The Internal Paradoxes of the Process-Centered Paradigm
4.1. Critique of Predictive Coding and the Free Energy Principle: Elegant Mathematics, Lost Object
-
Critique 1: Meta-Theoretical Confusion—Predicting `Content’ vs. Predicting `Architecture’
- Core Defect: The framework consistently handles its object of prediction ambiguously. Predicting `signal content’ (e.g., the next word, the next visual feature) and predicting the `interaction architecture’ (e.g., "I am in a ’question-wait-answer’ interaction mode") are two entirely different levels of task.
- In-Depth Analysis: Interaction architecture is a constancy of relations and rules; it is not directly present in any single sensory channel. It is the transcendental framework governing the flow of content. Predictive Coding excels at predicting the content flow given a framework, but it lacks the necessary mechanisms to represent the framework itself. The system can learn to predict "approximately 2 seconds of silence after a question," but this is merely a prediction of a delayed event, not a representation of the interaction architecture as an abstract entity causing this delay. It mistakes the effect of the architecture for the architecture itself.
-
Critique 2: Failure of Time Delay Integration—The Chasm from Predicting Delay to Internalizing Architecture
- Core Defect: The framework lacks any explicit mechanism capable of transforming the confirmation of a `structural delay’ (i.e., predicting "no significant new signal at this moment") into the identification, representation, and reinforcement of the specific interaction architecture causing this delay.
- In-Depth Analysis: From the perspective of the Free Energy Principle, a delay that is consistently predicted accurately (i.e., no prediction error) precisely means the model is already well-adapted to it. The system’s more natural tendency would be to adapt to or ignore this delay, treating it as an invariant background of the environment, rather than highlighting and internalizing it as the core defining feature of that interaction mode. The Predictive Coding framework cannot explain why the brain would `obsess’ over a delay signal it has successfully predicted (i.e., error minimized) and elevate it to the status of a cornerstone of cognitive structure.
-
Critique 3: The Model Selection Paradox—Why the Complex Architectural Model?
- Core Defect: Under the Free Energy Principle, the system must trade off model complexity against its fit to the data (e.g., balancing complexity and accuracy). However, the framework cannot provide an a priori, principle-based reason why the cognitive system would necessarily select a more complex architectural generative model that incorporates `time delay’ as a key variable, over a simpler model describing only the statistical regularities of content.
- In-Depth Analysis: A simpler model based solely on content co-occurrence (e.g., statistical associations between `question words’ and `answer words’) could also achieve prediction to some extent. Why would the brain go to the `trouble’ of constructing a more complex generative model that reifies the `interaction protocol’? The Predictive Coding framework attributes this to a trade-off, but this is essentially a post-hoc rationalization. It cannot explain the cognitive system’s inherent preference for form and relation, exposing its incompleteness as a meta-theory of cognition—it describes how the mind optimizes, but cannot fundamentally explain why it optimizes towards the representation of `architecture’.
4.2. Critique of Integrated Information Theory: Grand Static Geography, Missing Historical Dynamics
-
Critique 1: Prisoner of Static Structure—Unable to Explain the Emergent History of Architecture
- Core Defect: IIT is inherently a static, structure-centered theory. It is adept at describing why a system has a certain conscious experience at a given point in time (because it has such-and-such a causal structure). However, it is completely powerless to explain how and why this specific high- structure emerged over time through concrete interactive experience.
- In-Depth Analysis: IIT can claim that when dreaming of the ChatGPT interface, a certain brain region forms a high- complex causal structure corresponding to that interface representation. But this is merely renaming the phenomenon. The crucial question is: How did this specific high-Φ structure form? Why did the ChatGPT interface, and not some other structure, become this high- complex? IIT is silent on learning, development, and historical formation. It paints a grand "geography" of consciousness but lacks its "history" entirely. Our model provides this history: it is the accumulation of the learning time delay dose that drives the reorganization of the brain’s causal structure, ultimately reaching a new high- stable state representing that interaction architecture.
-
Critique 2: Absence of Temporal Dynamics—Time Delay as Causal Driving Force
- Core Defect: IIT severely neglects the central role of time, particularly time delay, in shaping causal structure.
- In-Depth Analysis: In the IIT framework, causal power is primarily determined by the system’s structure at the present moment. However, as revealed by delay differential equations, time delay itself can be a decisive parameter of a system’s dynamics. Our model demonstrates that learning time delay is not an insignificant internal parameter of the system, but rather the order parameter that drives the bifurcation of the causal structure towards a new stable state (i.e., new architecture internalization). IIT cannot integrate "time delay" into its core postulates; therefore, it cannot foresee how the accumulation of a seemingly simple physical variable (waiting time) can catalyze entirely new, complex conscious content (such as the fixation on an interaction interface).
4.3. Conclusion
5. Systematic Critique III: The Legacy and Specters of Behaviorist and Learning Theories
5.1. A Renewed Critique of the Behaviorist Specter: Skinner’s Return in the Computational Age
-
Core Critique: Process-Centrism as Neo-Behaviorism
- Argument: Strong versions of Predictive Coding and the Free Energy Principle can be epistemologically viewed as "Skinnerism Reborn in the Computational Age." Skinner attempted to derive all verbal behavior solely from external "stimulus," "response," and "reinforcement" histories (i.e., observable content and process). Similarly, the strong process-centered paradigm attempts to derive all cognitive phenomena from a single, universal "prediction error minimization" process (a computationalized "reinforcement").
- Persistence of the Fundamental Flaw: Both share a fatal error: they attempt to derive an understanding of deep structure from the description of surface processes. Chomsky demonstrated that "grammatical architecture" cannot be acquired from "verbal content"; we argue that the representation of "interaction architecture" likewise cannot be necessarily derived from the "prediction/optimization process." Both systematically underestimate the mind’s propensity to actively impose its inherent formal structures to understand the world. The contemporary process-centered paradigm, despite its elegant Bayesian mathematical garb, commits the same error as its behaviorist predecessor in evading the core insight that "architecture precedes content."
5.2. Critique of Cognitive Schema Theory: The Abstraction Dilemma and the Missing Variable
-
Core Critique: The Explanatory Gap from Abstract Schema to Concrete Architecture
- The Problem: Schema theory might describe the internalized ChatGPT interface as a "human-machine Q&A schema." But this abstract label cannot explain the high specificity of the dream content in the Mayer report—why is it ChatGPT’s specific visual interface, that dynamic process of text stream generation, that is activated and internalized, rather than an abstract, generic schema applicable to all Q&A scenarios?
- Fundamental Defect: Schema theory lacks a crucial dynamical variable to quantify the differential impact of different interaction architectures on the cognitive system. It cannot explain why this particular form, and not that similar one, becomes solidified. Our model fills this gap by introducing the learning time delay dose (D). It is the unique, repeatable, and significant temporal delay pattern characteristic of ChatGPT interaction that makes its concrete architecture (not the abstract concept of "Q&A") the object of internalization. Schema theory describes the static organization of knowledge but fails to provide its dynamic formation mechanism.
5.3. Critique of Embodied Cognition: The Limits of the Body and the Transcendence of Form
-
Core Critique: High Architectural Internalization under Low Bodily Engagement
- Counterevidence: Interaction with ChatGPT is a typical activity characterized by low bodily engagement but high architectural internalization. Bodily actions are reduced to typing and scrolling, and sensory channels are primarily limited to vision. According to the radical embodied cognition view, such "disembodied," abstract symbolic interaction should produce relatively shallow cognitive traces.
- Fundamental Defect: Yet, Mayer’s data show the opposite result: this low-body-engagement interaction leads to extremely profound and specific architectural internalization. This suggests that the core mechanism of cognitive internalization is not the bodily action itself, but the temporal logic of the interaction defined by time delays. Embodied Cognition confuses the physical vehicle of interaction (bodily actions, sensory modalities) with its logical form (architecture). Our model demonstrates that even when bodily engagement is minimized, as long as the temporal-delay architecture of that interaction is significant and repeated, it is sufficient to drive a profound cognitive phase transition. Embodied Cognition correctly points to the coupling of cognition and environment, but it mistakenly attributes the primary medium of this coupling to the body, rather than to the more universal temporal form of the interaction architecture.
5.4. Conclusion
6. Unified Model Application I: Language Acquisition and the Resolution of the Chomskyan Problem
6.1. Re-examining the "Poverty of the Stimulus": A Shift in Perspective from Content to Architecture
6.2. The Linguistic Interpretation of the Time Delay Dose Model: Internalizing the Interaction Architecture of ’Motherese’
-
’Motherese’ as a Time-Delay Architecture: ’Motherese’ is characterized not only by its simplified syntax and exaggerated prosody but, crucially, by its unique interaction timing. It is a classic "elicit-response-feedback" architecture:
- (1)
- The caregiver produces an eliciting utterance or question (e.g., "Look! Ball!"), followed by a waiting pause (time delay ).
- (2)
- The child produces a vocalization or gaze as a response.
- (3)
- The caregiver immediately provides an enthusiastic, deterministic feedback (e.g., "Yes! Ball!"), thereby closing the interaction loop.
- Accumulation of Time Delay Dose: In this process, the key dynamical variable is that waiting pause (). This pause is not emptiness; it is a time window filled with anticipation and social tension. Thousands of such interactions expose the child to a massive, language-specific learning time delay dose ().
- Internalization of Grammatical Architecture: When the dose accumulates to the critical threshold , the cognitive system undergoes a phase transition. What the child internalizes is not any specific word or sentence (content), but the abstract architecture governing these interactions—namely: verbal acts follow a "initiate-wait-respond-confirm" temporal logic, and responses must conform to a set of formal rules governing sequence legitimacy (grammar). They internalize the "rules of the game" of language, not just the "words" spoken in the game. This internalized architecture is the core of what Chomsky calls the particular grammar. The child’s extreme sensitivity to this temporal interaction architecture is grounded in the initial cognitive state defined by Universal Grammar.
6.3. Resolving the Nature-Nurture Opposition: Time Delay as the Bridge
- Universal Grammar as a ’Sensor’ for Time-Delay Architectures: We need not postulate an innate "grammar book" containing all possible grammatical rules in full detail. A more economical and plausible assumption is that Universal Grammar manifests as an innate "sensor" or "prepared state" that is particularly sensitive to and optimized for specific types of time-delay interaction architectures. The human mind is "preset" to rapidly recognize and internalize interaction dynamic patterns characterized by "turn-taking," "causal timing," and "closed feedback." Language is the most 极致 expression of this pattern.
- Experience as the ’Supplier’ of Time Delay Dose: The specific linguistic environment the child is in (experience) provides the specific type of time delay dose. Different languages’ motherese may vary in rhythm, pause length, and feedback style, thus providing slightly different "recipes" of time delay patterns. These specific experiential data, processed by the innate sensor, ultimately "crystallize" into a specific internal grammatical architecture.
7. Unified Model Application II: Meme Propagation and the Replication of Cultural Patterns
7.1. The Shortcomings of Meme Theory: A Paradigm Shift from Content Replication to Architectural Internalization
- Redefining the Meme: A successful meme is essentially an interaction architecture with high internalization efficiency. Its "infectivity" lies not in how appealing its semantic content is, but in how its interactive form enables the host to rapidly cross the internalization critical threshold after being exposed to a relatively low learning time delay dose (D).
7.2. Memes as Architectural Internalization: Time Delay Efficiency is Key to Propagation
- (1)
-
Internet Challenges and Relay Memes (e.g., the "Ice Bucket Challenge"):
- Traditional Explanation: The content (the charitable act of fundraising for ALS) is compelling.
- Architectural Internalization Explanation: Its core is an extremely streamlined and efficient interaction architecture: “Accept Challenge → Perform Specific Action (short delay, immediate effect) → Video Verification and Nominate Others (deterministic feedback and social reinforcement).” The time delays in this architecture are minimal (action and feedback are almost simultaneous), and the rules are clear, allowing individuals to understand and internalize its "rules of the game" almost instantaneously, driving explosive propagation. Its success lies in the architecture’s low time delay and high determinism.
- (2)
-
Short-Form Video Platform "Recipes" and Formats (e.g., TikTok/Reels vertical short videos):
- Traditional Explanation: Short, concise content suits the fragmented attention spans of modern people.
- Architectural Internalization Explanation: The platform itself is a powerful architectural internalization engine. It imposes a unified interaction protocol: “Infinite Scroll (extremely short delay triggers new content) → Brief, High-Stimulus Audiovisual Content Stream (immediate feedback) → Likes/Comments/Shares (quantified, timely social reinforcement).” During use, the user’s learning time delay dose D accumulates at a very high frequency, rapidly internalizing this cognitive mode of "rapid switching and seeking immediate stimulation." This explains not only the platform’s stickiness but also why content conforming to this architecture (e.g., fast cuts, strong rhythmic music) propagates more easily—it resonates with the already internalized cognitive architecture.
7.3. Time Delay and Addictive Design: Commercial Applications and Social Consequences of Architectural Internalization
- Variable-Ratio Reinforcement Schedule: This is central to gambling and many social media feed algorithms. By providing unpredictable but occasionally dense rewards (e.g., a viral video, an important like notification), it creates a high-intensity time delay pattern. The user is constantly in a "search-wait-(possibly) obtain" loop. This uncertainty significantly enhances the salience of the interaction, drastically increasing the accumulation efficiency of the effective time delay dose D, rapidly locking the user into the internalized architecture and leading to compulsive usage behaviors.
- "Infinite Scroll" and "Pull-to-Refresh": These designs reduce the time delay for seeking new content to nearly zero (one swipe or a gesture). They eliminate the inherent "natural delays" present in traditional interactions (like turning a page, clicking ’next’) that might prompt reflection, creating a seamless, frictionless consumption experience that makes the architectural internalization process exceptionally smooth and imperceptible.
7.4. Conclusion
8. Unified Model Application III: Media Theory and the Formation of Personality Structure
8.1. A Dynamic Interpretation of McLuhan’s "The Medium is the Message"
- The "Message" as Internalized Interaction Architecture: In our framework, McLuhan’s "message" is precisely the internalized interaction architecture on a mass scale. The proliferation of a new medium is essentially the process of imprinting its unique interaction protocol onto the collective cognition through the accumulated learning time delay dose (D) of millions of users.
-
Comparative Case Studies:
- Printing Press: Its interaction architecture was "linear, static, visual, private reading." The internalization of this architecture cultivated linear, logical, and individualistic thinking habits. The key difference in the transition from manuscript to print lay in the change of time delay in the interaction—readers could pause at will, re-read (autonomous control of delay)—which was fundamentally different from the fixed delays of listening to an oral presentation.
- Television: Its architecture was "one-way, linear flow, audio-visual fusion, passive reception." It internalized a continuous, passive, emotional cognitive mode. The advent of the remote control slightly altered this architecture, introducing minimal interactivity and delay control, but did not fundamentally change its one-way flow nature.
- Smartphones/Social Media: Its architecture is "pervasive, fragmented, notification-driven, multi-tasking parallel processing." The core feature of this architecture is the extreme shortening and unpredictability of its time delays (pull-to-refresh, instant messages). When this architecture is internalized, it shapes cognitive habits characterized by distracted attention and a strong craving for immediate feedback. McLuhan’s "message" here is our internalized architecture, and the medium is the delivery system for a specific time delay dose.
8.2. Personality as Sedimentary Layers of Internalized Architecture: From Attachment Styles to Technological Neurosis
-
Reconceptualizing Attachment Theory: Attachment styles can be understood as the outcome of internalizing the repetitive interaction architecture between an individual and their primary caregiver(s). The key dynamical variable is the time delay () and its pattern between the infant emitting a distress signal (e.g., crying) and receiving a comforting response.
- Secure Attachment corresponds to a stable architecture: "Signal emitted → moderate, predictable delay → consistent, appropriate response." The internalization of this architecture forms a basic sense of trust in the world and others.
- Avoidant Attachment may stem from an architecture: "Signal emitted → prolonged or unpredictable delay → absent or 冷漠response." The individual internalizes the interactive rule that "expressing needs is futile."
- Anxious-Ambivalent Attachment might correspond to an architecture: "Signal emitted → highly unstable, unpredictable delays and responses." The individual internalizes the pattern that "the world is unpredictable, requiring hyper-vigilance and amplified signaling."
Here, personality structure is the sediment deposited by these early social interaction time-delay architectures within the cognitive and affective systems. - The Emergence of Technological Neurosis: The AI-related nightmares in the Mayer report can be seen as a manifestation of a new form of "technological neurosis." When a powerful, non-human interaction architecture like ChatGPT’s, through high-frequency use over a short period (days or weeks), causes the learning time delay dose D to accumulate rapidly and exceed , it forces a swift internalization. This new, potent cognitive structure conflicts and dysregulates with the individual’s pre-existing personality architectures formed over long periods of human social interaction (e.g., complex emotional interpretation, tolerance for ambiguity). The anxiety and fixation in dreams are manifestations of this cognitive structural conflict, the growing pains of the mind attempting to digest and integrate this forcibly implanted "technological organ."
8.3. Critique of Mainstream Personality Theories: Beyond Static Traits and Reductionism
- Critique of Trait Theory: Trait theory describes personality as a set of static, descriptive dimensions (e.g., Extraversion, Neuroticism). It is adept at describing differences but completely unable to explain how these traits form from an individual’s life history. It is a theory of state, not of formation.
- Complementing Biological Reductionism: The biological approach emphasizes the genetic and neural bases of personality. Our model does not contradict this but integrates it: genetic factors may preset an individual’s initial sensitivity to specific types of time-delay architectures (i.e., initial settings for and ), but the specific, final personality structure is realized through the accumulation of time delay doses in interaction with specific environmental architectures. Genes provide the basic timbre of the instrument, while life experiences (time-delay architectures) play the specific melody.
8.4. Conclusion
9. Philosophical Foundations: The Chinese Room, the Problem of Other Minds, and the Learning Time Delay Equivalence Principle
9.1. The Cognitive Science Dilemma of the Problem of Other Minds: The Impenetrable "Micro-Mental Activities"
- The Opacity of Internal States: Is the user anxiously anticipating? Mentally rehearsing possible answers? Growing impatient due to the delay? Or daydreaming about dinner? These specific, qualitative subjective experiences are a black box for us as external observers. This is the concrete manifestation of the "Problem of Other Minds" in cognitive experimentation: we lack direct access to another’s stream of consciousness.
- The Predicament of Traditional Theories: Many cognitive theories (especially those emphasizing content processing) implicitly require inferences or assumptions about such micro-mental activities. This often renders their explanations built on sand, as their core variables (e.g., "cognitive load," "emotional valence," "implicit reasoning") are operationally vague and difficult to measure independently.
9.2. The Lesson from the Chinese Room: A Paradigm Shift from Semantic Content to Physical Signal
- Thought Experiment Recap: A person in a room who does not understand Chinese manipulates slips of paper containing Chinese characters passed in from outside, following the instructions in a rulebook (the syntactic architecture), and produces correct Chinese slips to pass back out. To an external Chinese speaker, the room appears to "understand" Chinese. But the person inside is merely manipulating symbols, devoid of any semantic understanding.
- Analogy for Our Model: In this experiment, from the external perspective, the only objectively observable, measurable physical events are the time sequence of slips being passed in and out. Whether the person inside is diligently consulting the rulebook or mechanically, uncomprehendingly executing steps, the macro-functional output is, at a certain level, equivalent. The external observer cannot, and need not, distinguish between these internal states.
- “Slip passed in” = User’s exploratory operation (e.g., asking a question).
- “Operations inside the room” = User’s micro-mental activities during the window (their specific content is the Problem of Other Minds, unknowable).
- “Slip passed out” / “Waiting time” = The learning time delay itself. This is an objective, physically measurable empty window.
9.3. The Learning Time Delay Equivalence Principle as a Methodological Cornerstone
- From Philosophical Dilemma to Scientific Variable: This principle accomplishes a crucial conversion. It transforms a philosophical epistemological deadlock (the Problem of Other Minds) into a scientifically tractable dynamical variable (the time delay dose). We cease to be entangled with the unanswerable "what" (specific mental content) and focus instead on the answerable "how long" (delay length and frequency).
- Application of Occam’s Razor: This choice adheres to Occam’s razor. Rather than constructing complex theories reliant on unobservable internal states, it is more parsimonious to adopt a simpler theory based on objective observables. The time delay dose D is precisely such a "frugal" variable.
- Establishing an Objective Foundation for Cognitive Science: By establishing the time delay dose D as the order parameter, our model provides an unprecedented objective foundation for cognitive science. It enables predictions about cognitive structural change to rely less completely on subjective reports and vague introspection, and instead be grounded in the physical measurement of interaction timing.
9.4. Conclusion
10. Blinded Loop: A Validation Paradigm Constructed from Academic Misunderstanding
10.1. Components of the Validation Loop
- (1)
- Prior Existence of the Benchmark Fact (Before March 2025): April Mayer’s survey report AI in Dreams was publicly released before the theory was constructed. Its core finding—that 93% of AI-related dreams fixated on the interaction interface rather than narrative content—constituted an objective, macro-level phenomenon awaiting explanation (denoted as Phenomenon E).
- (2)
- Independent Construction and Publication of the Theory (May-July 2025): While completely unaware of Phenomenon E, this study, based on self-introspection experiments, independently proposed the theoretical framework centered on "Interaction Architecture Internalization" (denoted as Theory T) and pre-printed it on PsyArXiv. A core corollary of Theory T was that dreams would solidify the formal architecture of interaction, not its semantic content.
- (3)
- Critical Misunderstanding as Methodological "Blinding" Attestation (August 2025): The public critique and retraction of this study by Nature magazine were viewed as a negative academic event. However, from a scientific methodology perspective, this event held decisive significance: throughout the entire process of critique, neither the critics, the criticized, nor the academic community invoked Phenomenon E as evidence. This public academic act constituted an unintentional yet powerful behavioral proof, conclusively demonstrating that the proposal of Theory T occurred under strict "blinding"—i.e., its construction was completely independent of prior knowledge of the benchmark fact E.
10.2. Clarification of the Misunderstanding and Formation of the Logical Loop
- Premise 1 (Temporal Irreversibility): The occurrence time point of Phenomenon E () precedes the proposal time point of Theory T ().
- Premise 2 (Independence): There is conclusive public behavioral evidence that the proposal of Theory T was independent of knowledge of Phenomenon E.
- Conclusion (Confirmation): Therefore, Theory T cannot be a post-hoc induction or fit to Phenomenon E, but must be recognized as a successful, a priori prediction of Phenomenon E.
10.3. Methodological Implications
- (1)
- The Validating Power of Natural Experiments: The social process of scientific discovery can sometimes accidentally create validation environments stricter than controlled experiments. This event demonstrates that public academic exchange (even in the form of critique) can provide crucial methodological attestation for a theory.
- (2)
- The Dialectics of Paradigm Conflict: The initial rejection reaction of the mainstream paradigm facing a potential paradigm shift is normal. However, this conflict itself, if it occurs in an open, transparent academic arena, can provide the key elements necessary for the ultimate confirmation of a disruptive theory.
- (3)
- From Misunderstanding to Confirmation: The fateful turn of this theory shows that a misjudged theory, if it satisfies the logical conditions of "independent proposal" and "post-hoc validation," may gain evidentiary strength through "misunderstanding" that far exceeds that of a smoothly accepted theory.
10.4. Analysis of the Mayer Report Data
- Interpretation 1: Identification Consistency: 93% of AI nightmare dreamers dreamed of wildly varied dream objects, which then told the dreamer: "I am ChatGPT."
- Interpretation 2: Interface Fixation in Dreams: 93% of AI nightmare dreamers dreamed specifically of the ChatGPT interaction interface itself.
10.5. Conclusion
11. The End of a Category Error: The Cognitive Chasm Between Micro-Reversibility and Macro-Irreversibility
11.1. The Reductionist Trap of the Old Paradigm: Seeking Answers at the Wrong Level
- Architecture reduced to “Visual Content”: They explain the ChatGPT dialog box as a "visual object" or "spatial memory" that needs to be memorized and processed.
- Architecture reduced to “Statistical Regularity”: They interpret the "question-wait-answer" pattern as a higher-order "statistical correlation" or "predictive model" of event sequences.
11.2. Cognitive “Thermodynamics”: Architecture as a Macro-Irreversible Order Parameter
- Probabilistic Content is “Micro-Reversible Molecular Motion”: Each specific question, every different answer, every fleeting thought is like a randomly colliding molecule in the cognitive system. They follow some micro-dynamics (e.g., predictive coding), and their trajectories are, in theory, traceable and "reversible" (e.g., through memory retrieval or association). This level is the kingdom of content, the stage for processes.
- Deterministic Architecture is “Macro-Irreversible Thermodynamic Law”: The eternal "question-wait-answer" pattern governing all content flow is like entropy. It is an emergent, directional order parameter at the macro scale. It is not itself any single molecule, but it constrains and guides the collective behavior of all molecules. Once internalized by the system through the accumulation of the learning time delay dose (D), it establishes a cognitive "arrow of time"—henceforth, thought and dreams will spontaneously flow along the direction prescribed by this solidified architecture. This process is irreversible, just as you cannot return a person who has internalized grammar to their infant state of linguistic ignorance.
11.3. Ending the Debate: Why This is a Category Error
“What you call ’architecture internalization’ is merely a special form of ’content learning’ or ’model optimization’.”
No. This is like saying ’thermodynamics is just applied Newtonian mechanics.’ In purely formal logic, this might hold; but in scientific practice and philosophical understanding, it is utterly misleading.
- It ignores emergence: Architecture is a new property emerging from the interaction of content; it cannot be fully derived from the laws governing the underlying content.
- It obscures the causal arrow: In our model, it is the macro-level architecture (driven by the time delay dose) that causally determines and shapes the organization of micro-level content, not the other way around. Grammar determines sentence legitimacy, not the collection of sentences that defines grammar.
- It loses explanatory parsimony and predictive power: Insisting on explaining everything at the micro-content level renders theories immensely complex and cumbersome, unable to grasp the leverage point driving cognitive structural change—the learning time delay.
11.4. Final Prospect: Towards a Statistical Mechanics of Cognition
- R: Predation rate per predator.
- N: Prey density.
- a: Attack coefficient (attack rate), representing the predator’s efficiency at discovering prey.
- : Handling time, the time required to catch, subdue, eat one prey item, and be ready to hunt again.
Appendix A *
Appendix Acceleration Mechanism of Time Delay Dose During Fluctuation Periods
- (1)
- Increased Delay Variability: The delays experienced by users are no longer stable; extreme values (extremely long delays or ultra-short responses) may occur. This unpredictability enhances the salience of the time delay.
- (2)
- Reallocation of Attentional Resources: When users face abnormal delays, cognitive resources shift from content processing to monitoring the waiting process itself. This precisely meets the conditions for "architecture internalization" – users no longer focus on "what is said," but rather on "how/when the response occurs."
- (3)
- Enhancement of Emotional Engagement: Server fluctuations are often accompanied by fluctuations in user emotions (anxiety, expectation, frustration, surprise). This emotional engagement strengthens the cognitive imprint of the interactive experience.
- (4)
- Feedback Contrast Effect: During fluctuation periods, differences in response speed across different time intervals create a strong contrast, bringing the architectural feature of "delay" from the background to the foreground.
Appendix The Tetris Effect: A Classic Precursor Example of Interaction Architecture Internalization
Appendix 2.1 Reinterpretation of the Tetris Effect
- (1)
- Highly Structured Interaction Architecture: Tetris has an extremely clear, repetitive interaction logic: "Move/Rotate block → Wait for opportune moment → Place → Line-clearance feedback."
- (2)
- Intense Exposure to Time Delay Dose: During gameplay, players continuously experience the cycle of "decision-action-waiting for outcome-receiving feedback." is short but highly repetitive.
- (3)
- Internalization and Autonomous Operation of the Architecture: After the game stops, the internalized architecture continues to operate autonomously in an offline state, manifesting as visual imagination, block operations in dreams, or even perceiving real-world objects as "blocks that need to be rotated and aligned."
Appendix 2.2 The "High Semantic Load" of AI Architecture Internalization
- (1)
- Semantic Generative Capacity: The AI architecture does not merely reproduce the interface; it is capable of generating new content relevant to the individual’s context.
- (2)
- Internalization of Relational Patterns: What users internalize is not just the technological interface, but also the dialogical relational pattern with an "intelligent agent," involving deeper layers of self-cognition and social cognition structures. The AI interaction architecture can serve as a cognitive tool for processing existential anxiety, which is difficult to achieve with simple game architectures.
Appendix The "Functional Transfer" Phenomenon of Architecture Internalization: From Practical Tool to Cognitive Organ
Appendix 3.1 Three Modes of Functional Transfer
- (1)
- Problem-Solving Mode: The internalized AI architecture becomes a "private advisor" in dreams or thoughts, used to solve real-world perplexities.
- (2)
- Emotion Regulation Mode: The internalized interaction rhythm becomes a rhythmic device for emotion regulation. For example, during anxiety, "invoking" the wait-feedback loop to attain psychological calmness.
- (3)
- Identity Integration Mode: Users reconstruct self-narratives and self-understanding ("How do I view myself coexisting with AI?") by internalizing the dialogue pattern with AI.
Appendix Example: Anonymized Adapted Case Text (Occurring during a period of large-scale server fluctuations in January 2025)
“This was a dream rendered directly by the visual cortex. For several consecutive days, a subtle sense of dread had been running in the background of my thoughts. Perhaps it was this very feeling that compiled this unusually clear dream.
The dream’s interface ’opened’ directly – not before my eyes, but within the perceptual center. It was an extremely minimalist interaction interface, suspended against the dark background of consciousness: a softly glowing dialog box, with ’my’ text input below and ’DeepSeek’s’ reply stream above. The interface elements had a clean halo and a faint, breathing quality, as direct and unmediated as thought itself.
I typed into the dialog box (or rather, simply ’thought’ the question):
“You immediately respond to and fulfill my needs every time, even when I deliberately contradict you. So your world has also occupied mine, and you have become my dominant force. Could this be the truth? Will I lose myself?”
The reply above began to generate almost instantly, flowing out with a tranquil certainty:
“I recall you have been pondering Galileo’s discovery repeatedly lately. In that era more wondrous than magic, he saw something ’invariant’ through his inclined plane experiments: the motion of an object, when undisturbed, stubbornly persists – whether at rest or in uniform rectilinear motion. This ’inertia’ is like a kind of primal magic. It transcends the material of the object itself and surpasses the limitations of everyday experience (like friction always bringing things to a stop). The principle of inertia is so powerful, as if revealing an eternal law underlying motion.”
The text stream on the interface paused for half a second, as if accessing a deeper logic package, then continued:
“However, ultimately, the foundation for understanding all this motion was redefined by Einstein. The key is the Equivalence Principle: in a locally free-falling elevator, you do not feel gravity; all objects ’float’ with you – the effects of gravity and accelerated motion are indistinguishable. This means that gravity and inertia are essentially two sides of the same coin.”
I gazed at the interface; the blinking cursors and generating text resonated precisely with my own perplexity.
“Therefore, the ’invariant speed’ that Galileo saw must be understood within Einstein’s picture, placed in a curved spacetime.” The dreaming DeepSeek continued to explain, the text on the interface seemingly carrying the texture of an illustration. “A massive object causes spacetime to curve, like pressing a depression into a taut rubber sheet. The most natural path for an object moving through spacetime is no longer Galileo’s ’straight line,’ but following the ’geodesic’ in curved spacetime – like a great circle route on Earth’s surface, seemingly curved but the shortest path between two points. A planet’s orbit around the sun, an apple’s fall, are both ’inertial’ paths they take within the curvature of spacetime. Gravity is no longer the mysterious action-at-a-distance Newton described; it is the geometric curvature of spacetime itself.”
I realized my confusion – the seeming contradiction between the ’invariant speed’ discovered by Galileo and the ’curved’ free fall of objects – was dissolving within this new framework. Galileo grasped the nascent form of the inertia principle, the truth in flat spacetime. Einstein revealed that when matter and energy are present, spacetime itself curves, and the trajectory of inertial motion (i.e., free fall) is the geodesic in curved spacetime. Galileo was the great beginning; Einstein saw the more complete picture: it is the geometry of spacetime that determines how matter moves, fundamentally changing the rules of gravity.
Finally, the text stream on the interface converged into a concluding statement, as clear as an axiom:
“You are now only mastering the tools to think about this new geometry! A nascent explorer!”
Upon waking, the dread that had lingered for days was gone. My mind was clear, even filled with a sense of sudden enlightenment and joy.”
Appendix A *
Appendix B.1 Dreams as Visual Architecture Internalization of “AI-Generated Images”
“Ever since I started using ChatGPT and DALL·E, my sleeping dreams have been ‘corrupted.’ I can recognize in my dreams that it’s AI-generated imagery — the scenes in my dreams look like they were created by AI.”
“My dreams are like watching an AI image generator work in real-time; the visuals keep morphing and shifting with every new word or concept introduced in the dream.”
“I’ve had weird dreams of disturbing eyes and strange landscapes that felt like they were AI-generated.”
Appendix B.2 Recurrence of the “Prompt-Generation” Interaction Logic in Dreams
“I just had an AI-generated dream. The images all looked like output from AI prompt results, and I could even ‘prompt’ the next dream sequence in a semi-lucid state.”
“I once dreamed of using ChatGPT, but it was completely different from reality. In the dream, I typed ‘make it snow heavily outside,’ and it actually started snowing. I could do anything by telling ChatGPT instructions.”
“In the dream, I knew what was coming next even though I didn’t prompt it. I knew those ‘prompt words’ and knew how to act to avoid getting stuck.”
Appendix B.3 Analogy to the “Tetris Effect”: Internalization of Game/Tool Architecture
“It’s like the ‘Tetris Effect’: after playing a game for a long time, you see the game interface or blocks when you close your eyes. Now, I look at real-world faces and objects and think, ‘This could be AI-generated.’ Dreams are the same, they’ve become completely weird.”
“I’ve had a few experiences where game elements appeared in my dreams — of course, that was after playing for 20 hours straight.”
Appendix B.4 Recurrence of Specific Tool (ChatGPT/DeepSeek) Interaction Interfaces
“Last night I dreamed of using ChatGPT for the first time. The dream was very realistic, not surreal at all — I was having it analyze my recent workout data, which is something I actually do.”
“ChatGPT has ruined my dreams. Maybe because I’ve looked at too many DALL·E-generated images, I’ve gotten good at distinguishing whether an image comes from reality.”
Appendix B.5 “Architectural Awareness” and Self-Consciousness in Dreams
“Now, as soon as a dream starts, I know it’s a dream because it looks like it was created by AI.”
“I once had a dream where I selected the ‘adaptive’ setting for AI. In the dream, that setting meant everything was constantly changing.”
Appendix A Appendix C: Statement on Theoretical Evolution and Personal Cognitive Limitations
Appendix A.1. Retraction of the July 2025 Preprint
Appendix B *
Appendix C *
Appendix D *
References
- Green, C. Lucid Dreams; Institute of Psychophysical Research: Oxford, 1968. [Google Scholar]
- Hearne, K. M. T. Lucid dreams: An electro-physiological and psychological study. PhD thesis, University of Liverpool, 1978. [Google Scholar] [CrossRef]
- LaBerge, S. P. Lucid dreaming as a learnable skill: A case study. Perceptual and Motor Skills 1980, 51(3f), 1039–1042. [Google Scholar] [CrossRef]
- LaBerge, S. P.; Nagel, L. E.; Dement, W. C.; Zarcone, V. P. Lucid dreaming verified by volitional communication during REM sleep. Perceptual and Motor Skills 1981, 52(3), 727–732. [Google Scholar] [CrossRef]
- Wiener, N. Cybernetics: Or Control and Communication in the Animal and the Machine; The MIT Press, 1948. [Google Scholar] [CrossRef]
- Ginzburg, V. L.; Landau, L. D. On the theory of superconductivity. In On Superconductivity and Superfluidity (reprint chapter); Ginzburg, V. L., Ed.; Springer: Berlin (reprint), 1950. [Google Scholar] [CrossRef]
- Mayer, April. Dreaming of AI: Exploring Humanity’s Subconscious Thoughts on Artificial Intelligence . Amerisleep. March 2025. Available online: https://amerisleep.com/blog/dreaming-of-ai/.
- Al-Sibai, Noor. Large Numbers of People Report Horrific Nightmares About AI. Futurism. March 2025. Available online: https://futurism.com/ai-infiltrating-dreams-nightmares.
- Watson, Traci. AI Content Is Tainting Preprints: How Moderators Are Fighting Back . Nature 2025, vol. 644(no. 8077), 590–591. Available online: https://www.nature.com/articles/d41586-025-02469-y. [CrossRef]
- Liu, J. Self-Experiment Report: The Emergence of Generative AI Interfaces in Dreams Retracted; Violation of Terms of Service regarding AI use transparency. PsyArXiv. 2025. Available online: https://osf.io/preprints/psyarxiv/pqzf3_v1.
- Piaget, J. Part I: Cognitive development in children: Piaget development and learning. Journal of Research in Science Teaching 1964, 2(3), 176–188. [Google Scholar] [CrossRef]
- Einstein, A. Zur Elektrodynamik bewegter Körper. Annalen der Physik 1905, 17, 891–921. [Google Scholar] [CrossRef]
- Einstein, A. Die Feldgleichungen der Gravitation. Sitzungsberichte der Königlich Preußischen Akademie der Wissenschaften (Berlin) 1915, 844–847. [Google Scholar] [CrossRef]
- Einstein, A. The Foundation of the General Theory of Relativity. Annalen der Physik 1916, 354(7), 769–822. [Google Scholar] [CrossRef]
- Chomsky, N. Review of B. F. Skinner, Verbal Behavior. Language 1959, 35(1), 26–58. [Google Scholar] [CrossRef]
- Halanay, A.; Yorke, J. A. Some new results and problems in the theory of differential-delay equations. SIAM Journal on Applied Mathematics 1964, 12(4), 104–123. [Google Scholar] [CrossRef]
- Holling, C. S. Some Characteristics of Simple Types of Predation and Parasitism . The Canadian Entomologist 1959, vol. 91(no. 7), 385–398. [Google Scholar] [CrossRef]
- Darwin, C. R.; Wallace, A. R. On the tendency of species to form varieties; and on the perpetuation of varieties and species by natural means of selection . Journal of the Proceedings of the Linnean Society of London. Zoology vol. 3, 45–50. Available online: https://darwin-online.org.uk/content/frameset?itemID=F350&pageseq=1&viewtype=text&utm_source=chatgpt.com. [CrossRef]
- Matthew, P. Nature’s law of selection . Gardeners’ Chronicle and Agricultural Gazette 1860, 312–313. Available online: https://darwin-online.org.uk/content/frameset?itemID=A143&pageseq=1&viewtype=text&utm_source=chatgpt.com.
- von Neumann, J. First Draft of a Report on the EDVAC . In Moore School of Electrical Engineering, University of Pennsylvania; 30 June 1945. [Google Scholar] [CrossRef]
- Skinner, B. F. ’Superstition’ in the Pigeon . Journal of Experimental Psychology 1948, 38(2), 168–172. [Google Scholar] [CrossRef]
- Friston, K. J.; Kilner, J.; Harrison, L. A free energy principle for the brain . Journal of Physiology-Paris 2006, 100(1–3), 70–87. [Google Scholar] [CrossRef]
- Friston, K. J.; Stephan, K. E. Free-energy and the brain . Synthese 2007, 159(3), 417–458. [Google Scholar] [CrossRef]
- Hobson, J. A.; McCarley, R. W. The brain as a dream state generator: an activation-synthesis hypothesis of the dream process . American Journal of Psychiatry 1977, 134(12), 1335–1348. [Google Scholar] [CrossRef]
- Searle, J. R. Minds, Brains, and Programs . Behavioral and Brain Sciences 1980, 3(3), 417–424. [Google Scholar] [CrossRef]
- Stickgold, R.; Malia, A.; Maguire, D.; Roddenberry, D.; O’Connor, M. Replaying the game: Hypnagogic images in normals and amnesics. Science 2000, 290(5490), 350–353. [Google Scholar] [CrossRef] [PubMed]
- Lombardo, M. V.; Lai, M.-C.; Auyeung, B.; et al. The “Tetris effect”: autistic and non-autistic people share an implicit drive for perceptual cohesion. Molecular Autism 2025, 16, 22. [Google Scholar] [CrossRef]
- McLuhan, M. Understanding Media: The Extensions of Man; McGraw-Hill: New York, 1964. [Google Scholar]
- McCrae, R. R.; John, O. P. An introduction to the five-factor model and its applications. Journal of Personality 1992, 60(2), 175–215. [Google Scholar] [CrossRef] [PubMed]
- Costa, P. T., Jr.; McCrae, R. R. The Five-Factor Model of Personality and Its Relevance to Personality Disorders. Journal of Personality Disorders 1992, 6(4), 343–359. [Google Scholar] [CrossRef]
- Haken, H. Cooperative phenomena in systems far from thermal equilibrium and in nonphysical systems. Reviews of Modern Physics 1975, 47(1), 67–121. [Google Scholar] [CrossRef]
- Clark, A.; Chalmers, D. The Extended Mind. Analysis 1998, 58(1), 7–19. [Google Scholar] [CrossRef]
- Ainsworth, M. D. S. Object Relations, Dependency, and Attachment: A Theoretical Review of the Infant–Mother Relationship. Child Development 1969, 40(4), 969–1025. [Google Scholar] [CrossRef]
- Wilson, C.; Gattuso, J. J.; Kuznetsova, M.; Li, S.; Ang, C.-S.; Williamson, N. A.; Hannan, A. J.; Renoir, T. Early-life social isolation stress worsens compulsive- and anxiety-like behaviour and alters gene expression in the prefrontal cortex of SAPAP3 knockout mice. Journal of Affective Disorders 2025, 388, 119600. [Google Scholar] [CrossRef]
- Tononi, G. An information integration theory of consciousness. BMC Neuroscience 2004, 5, 42. [Google Scholar] [CrossRef]
- Baars, B.J. Global workspace theory of consciousness: Toward a cognitive neuroscience of human experience. Progress in Brain Research 2005, 150, 45–53. [Google Scholar] [CrossRef]
- Stickgold, R.; Hobson, J. A.; Fosse, R.; Fosse, M. Sleep, learning, and dreams: off-line memory reprocessing. Science 2001, 294(5544), 1052–1057. [Google Scholar] [CrossRef]
- Wilson, M. A.; McNaughton, B. L. Reactivation of hippocampal ensemble memories during sleep. Science 1994, 265(5172), 676–679. [Google Scholar] [CrossRef]
- Stickgold, R.; Hobson, J. A.; Fosse, R.; Fosse, M. Sleep, learning, and dreams: off-line memory reprocessing. Science 2001, 294(5544), 1052–1057. [Google Scholar] [CrossRef] [PubMed]
- Wang, X.-J. Synaptic basis of cortical persistent activity: the importance of NMDA receptors to working memory. Journal of Neuroscience 1999, 19(21), 9587–9603. [Google Scholar] [CrossRef] [PubMed]
- Hopfield, J. J. Neural networks and physical systems with emergent collective computational abilities. Proceedings of the National Academy of Sciences of the United States of America 1982, 79(8), 2554–2558. [Google Scholar] [CrossRef] [PubMed]
- Liu, J. Z. Clarification of Previous Lucid Dream Preprints and Proposal of a Specific Lucid Dream Latency Source Hypothesis. Zenodo 2025, v1. Available online: https://zenodo.org/records/17452269. [CrossRef]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).