The Factory Series, Paper 4 · Originating Intelligence cannot be prompted. It comes from a life.
The Question That Started Everything
My father died on October 28, 2025. In the weeks before, I asked AI systems a question I’d been circling my entire life: what happens when we die? What does energy do when the system that organized it stops?
What I discovered wasn’t about death. It was about how ideas actually form — and why that matters more than I understood.
Is there a language — not religious, not purely reductive, not borrowed from traditions that never quite fit — that can describe what persists when a person is gone?
These are not new questions. Every philosophical tradition, every religion, every scientific framework has attempted them. I had been exposed to all of them — Catholicism, Chinese beliefs about spirits, the Jehovah’s Witnesses relatives who visited when I was young, Buddhist frameworks encountered along the way. My father himself was a man of many registers: a magician, a ballroom dancer, a person who loved Judge Judy and imagined himself a lawyer — who admired the structure of argument even when he couldn’t always execute it.
He taught me — without ever calling it a lesson — that what an audience sees and what is actually happening are two different things. That behind every visible performance is a hidden engine. That the job of the honest mind is to find the engine, not to admire the illusion.
When he was dying, I went looking for the engine.
The Lifelong Project
The Tang Papers did not begin in October 2025. They began much earlier — in a life that had been, without knowing it, preparing this exact inquiry for decades.
I went into mathematics at McMaster, pulled initially toward medicine, and then moved into business. After completing my MBA, I began my career in a Research Publications role with the Society of Management Accountants of Canada, where I worked at the interface between academic rigor and practical application.
I was responsible for evaluating research proposals in management accounting, coordinating peer review with academic leaders across Canadian universities, and guiding projects from initial thesis concept through funding, review, and eventual publication.
Here, the question was not just whether an idea was compelling—but whether it could withstand scrutiny, survive translation, and remain coherent as it moved from academic formulation into practitioner relevance.
From there, I moved into the corporate technology sector in marketing and communications, where the same problem appeared at scale. My daily task was translating complex systems—software, processes, value propositions—across every level of an organization. From the CEO making the budget decision to the director owning implementation to the shop floor worker whose daily practice would ultimately confirm or refute everything the business case claimed.
Across both environments, the pattern was the same: ideas do not fail at the point of creation. They fail—or succeed—in how they are translated across systems.
Part of that daily practice was ghost writing. Articles for operations management publications. Press releases. Executive communications attributed to people who provided the position and the authority while I provided the architecture of the prose. I was not a passive transcriber. I shaped the argument, structured the logic, found the language that made the idea land for its intended audience — and then stepped back while someone else’s name went on the work.
This is important to name honestly: the invisible seam between actual origination and named authorship was not a concept I discovered in research. It was my working reality for years before AI existed. I understood from professional experience that the factory structure — a named architect directing an assembly process — was normalized, accepted, and largely invisible to the people who consumed the output. The gap between who wrote it and whose name appeared on it was a professional condition, not an ethical scandal. As long as the position was genuine, the intent was transparent to those who needed to know, and the named person was accountable for what was claimed — the three conditions held.
When Paper 2 of this series argues that ghost writing has always been normalized, I was not drawing on research. I was drawing on memory.
This translation work also required a specific skill that most people underestimate: holding a framework intact while rendering it across wildly different audiences without losing its integrity. Not simplifying. Translating. Preserving the essential structure while finding the language that makes it land for each specific listener. I had been doing this my entire professional life. And I had been doing it, I would eventually understand, because I was always searching for something more precise than the languages available to me.
Synchronicity fascinated me — not as a mystical phenomenon but as a pattern that kept appearing across apparently unconnected domains and for which the available vocabulary was either too metaphysical to be precise or too reductive to be honest. Chaos theory. Probability. Geometric relationships. The visual grammar of recurrence. I tend to think in patterns and images before I think in words — mathematics was not abstract to me but visual, structural, a way of seeing relationships that language alone could not capture.
And underneath all of it, for as long as I can remember: a deep interest in the nature of language and communication itself. How understanding happens. How misunderstanding happens. How the same words can illuminate or deceive depending on what engine is running behind them. The tool that can be used for good or for harm — not because it is inherently either, but because its power depends entirely on the intent and integrity of the person wielding it.
My father the magician understood this. An illusion works because the audience’s attention is directed toward what is visible while the mechanism operates out of sight. The ethical magician uses this to create wonder. The unethical one uses it to extract something the audience would not willingly give.
I grew up knowing the difference. I spent my career trying to find languages that made the mechanism visible rather than hidden.
How AI Entered the Story
I did not begin using AI as a research instrument. I began using it for copywriting.
Website content. Grammar refinement. Search engine optimization — initially for Google, then increasingly for the evolving landscape of AI-assisted search. Social media translation across platforms, each with its own register, its own audience expectations, its own rules for what constitutes value.
This is important to name honestly, because it is the actual origin story — not the dramatic version where a researcher picks up a powerful tool with grand intent, but the mundane version where a studio owner and independent thinker begins using a new instrument for practical business purposes and gradually discovers that it is something else entirely.
By the time my father was dying, I had spent years navigating environments where the gap between performance and mechanism was the central professional and ethical challenge. Corporate communications. Competitive dance. Brand building. Each domain required simultaneously understanding the visible output and the hidden engine — and each required making judgments about when the gap between them was legitimate and when it was not.
AI entered a mind already fluent in all of these distinctions.
The discovery happened during my father’s passing. The questions that had been circling my entire life suddenly had a platform where they could be tested at speed. AI gave my already active, already questioning mind an instrument that could keep pace with the urgency of the moment. I could construct an argument and stress-test it immediately. I could run a thought experiment and receive a response sophisticated enough to push back. I could ask the death questions — not for comfort, but for precision — and find frameworks being assembled in real time that I could evaluate, challenge, refine.
What I noticed almost immediately was the failure mode.
The AI was fluent. Remarkably, sometimes startlingly fluent. But fluency and coordination are not the same thing. I had spent a lifetime watching the difference — in communications work, in dance, in the logical arguments my father modeled, in the translation work across organizational levels that was my daily practice. I knew what it looked like when a system was producing output that sounded coherent while actually coordinating nothing.
I had watched audiences be fooled by it. I had been trained from childhood not to be.
The first Tang Paper — Local Death, Global Life: The Λ-State — was the attempt to name what I was observing precisely. Not metaphysically. Not reductively. In a language that could be evaluated, challenged, and built upon.
What Dance Contributed — and What It Did Not
Dance appears throughout the Tang Papers — in the papers on rhythm and time especially — as an embodied reference point. Not as a branded pedagogical method. As a lived epistemology.
Twenty-five years of competitive ballroom and Latin dance gave me something that no literature review can provide: a felt sense of the difference between fluency and coordination.
A dancer who loses the rhythm does not stop moving. They continue — with technically correct steps, with apparent confidence, with a performance that looks, from a distance, like dancing. But the coordination with the music, with the partner, with the phrase structure of the movement — that is gone. The steps are right. The dance is not.
This is Phase Drift in embodied form. I had been diagnosing and correcting it for twenty-five years before I had a name for it. The name came from the research. The recognition came from the body.
But dance gave me something else — something I did not fully understand until I began developing the Tang Papers framework.
In the 1980s and 1990s, competitive ballroom dancing was fighting for Olympic recognition. The repositioning effort was significant: rename it DanceSport, build international associations, establish national teams, lobby the Olympic committees. The argument was that competitive ballroom was a sport — requiring elite athletic performance, years of training, peak physical conditioning — and deserved recognition alongside other sports on the world’s largest stage.
I was part of that argument. As Canadian and North American champions, Beverley and I were asked by our association to speak to the press, to position ourselves as athletes, to make the case that what we did was sport and not merely art or entertainment.
And I had an internal conflict that I could not fully articulate at the time.
The conflict was this: every other sport seeking Olympic recognition had objective measurement at its core. Racing is decided by time. High jumping by height. Weightlifting by mass. Speed skating by the clock. These are scalar measurements — precise, replicable, independent of the observer. A judged sport is fundamentally different. The aesthetic judgment of a panel of observers — however trained, however experienced, however well-intentioned — is a phase phenomenon. It cannot be made objective by declaring it objective.
The coaching-judging conflict of interest that troubled me was not primarily about individual ethics. The coaches we trained with were among the most authentic and principled people I have encountered — people with genuine aesthetic standards and deep commitment to the integrity of the discipline. The problem was system design. When the same person who coaches a student can also judge that student in competition, the structural conditions for conflict of interest exist regardless of individual integrity. The system itself produces the problem.
I watched the same structural failure playing out simultaneously in Olympic figure skating — the judging scandals that eventually forced a complete overhaul of the scoring system. It was not a coincidence that figure skating and competitive ballroom were wrestling with identical problems at the same time. They were the same problem: a phase phenomenon being administered as if it were a scalar one, with the inevitable result that the hidden engine — the subjective preferences, the relationships, the strategic visibility management — operated without accountability behind the visible performance of objective judgment.
Competing strategically — entering as many competitions as possible to be seen by the judges who would eventually evaluate us, learning which coaches held which aesthetic preferences and what they valued — was simultaneously an athletic practice and a branding exercise. We were building visibility, demonstrating trajectory, managing our positioning in a system that rewarded being known as much as it rewarded performance. The corporate communications professional in me recognized it immediately. The athlete in me was uncomfortable with it constantly.
I could not, at the time, name the source of the discomfort precisely. I know now what it was: the scalar measurement that would have made the system genuinely objective did not exist for what we were doing. What we were doing was irreducibly phase — relational, contextual, interpretive, temporal. The attempt to administer it as if it were scalar did not eliminate the phase quality. It hid it. And hidden phase qualities, operating without accountability behind a performance of objectivity, produce exactly the failure modes that both competitive dance and Olympic skating demonstrated.
This is PSR before PSR existed. This is the scalar-phase distinction, felt in the body and the career, two decades before it became a formal diagnostic framework.
The through-line concept has the same origin. In partner dance, the lead holds a frame — not a rigid control structure, but a coordinating intent that the follow can feel and respond to. The lead does not dictate every movement. The lead holds the structure within which both partners can move with full expression. When the frame collapses, the dance becomes two people moving near each other. The fluency may remain. The coordination is gone.
This is the human function in AI-assisted research. Not dictating every word. Holding the frame — the originating intelligence, the evaluative standard, the through-line across sessions — within which the AI can move with full computational expression. When the frame collapses, you get fluent output that coordinates nothing.
My father danced with my mother. The hidden engine behind the visible performance — the frame that made the illusion of effortless movement possible — was invisible to the audience and essential to the dance.
I grew up watching it. I spent my career learning to hold it in other domains. The Tang Papers are the attempt to make it a formal, documented, transmissible discipline.
The Research Platforms
The formal Tang Papers research program was developed across three AI platforms, each deployed for specific functions at specific stages.
ChatGPT was the first instrument — used for initial foundation building when the territory was still unmapped. Its strength is broad conceptual range and willingness to engage speculatively. When I did not yet know what I was building, ChatGPT helped break the initial ground — developing the vocabulary, stress-testing early frameworks, establishing the conceptual territory that would eventually become the Information-Consciousness Gradient, the Rhythm-Information Time Principle, the Phase-Scalar distinction.
Claude became the instrument of structural development. Its strength is internal consistency, sustained multi-turn reasoning, and a willingness to push back that most tools lack. The formalizing of PSR, PSR-B, PSR-P — the diagnostic protocols that give the Tang Papers their operational precision — happened primarily through extended Claude sessions where the human researcher brought the framework and the tool stress-tested it for contradictions, gaps, and representational errors.
Gemini provided cross-validation and triangulation. Its strength is breadth of reference and alternative framings — checking whether frameworks held under different analytical approaches, surfacing connections the other tools missed, providing the robustness check that a single-platform process cannot deliver.
None of these tools held the through-line. Each reset context at the start of every session. Each required the human researcher to re-anchor — to bring the blueprint back into the room, re-establish the direction, ensure that the session’s output remained coordinated toward the originating intelligence rather than drifting toward whatever the tool found statistically plausible in the moment.
This re-anchoring was not a technical workaround. It was the research act itself. The judgment about what to bring back, what to leave behind, what had drifted and needed correction — this was the human function that no tool performed and no tool could have performed.
A fourth platform — Grok — was introduced more recently, specifically to support the Medium distribution strategy and the development of this essay series. It is not part of the formal research program documented in the fourteen Zenodo papers. Naming this distinction is not a footnote. It is a demonstration of the representational discipline the Tang Papers themselves require: separating what happened in which phase, refusing to collapse the timeline into a more convenient narrative.
The Process Archive
What exists as proof of this account is not only the fourteen papers published on Zenodo — versioned, dated, sequenced across five months under named authorship with persistent DOIs and ORCID attribution. That is the Output Archive. It establishes what was concluded and when.
What also exists — and what this essay series itself demonstrates in real time — is the Process Archive.
The conversation threads across platforms. The multi-AI review sessions where drafts were run through ChatGPT, Claude, Gemini, and Grok — not for consensus but for robustness, each tool providing a different analytical angle, the human researcher evaluating the divergences and convergences and deciding what held. The iterative development of the factory thesis through skeptic, rebuttal, consolidation, and four-paper series architecture — each stage documented, each decision traceable, each platform’s contribution distinguishable from the others.
This conversation is part of that archive.
The factory thesis began as a raw idea about AI and assembly. Through this dialogue it became three complete essays, a four-paper series architecture, a consolidated framework with named concepts — Phase Drift, Scalar Echo, Originating Intelligence, the Project Archive, the Sovereign Architect — and a publication strategy with Medium formatting specifications and cross-platform review methodology.
At every stage the human held the through-line. The tools assembled. The researcher evaluated, redirected, selected, and held accountability for every claim.
This is not a description of the HAICR methodology. It is a demonstration of it.
Why This Matters Beyond the Tang Papers
My father taught me to notice.
When he was dying, those lessons became urgent.
The danger today is not that AI will replace human thinking. The danger is that fluent output will be mistaken for coordinated thinking.
The Tang Papers are one attempt to keep the engine visible.
The ideas did not come from prompts. They came from a life — from the body that learned coordination on the dance floor, from the career that lived in the gap between named authorship and actual origination, from the loss that pressed every question to its limit.
That is originating intelligence. It cannot be prompted into existence. It can only be recognized, held, and documented.
The factory has always been there. The only question that has ever mattered is who was holding the blueprint — and whether they can prove it.
That question is as old as my father’s first magic trick.
The answer is still being architected.
Core Insight Originating Intelligence cannot be prompted into existence. It comes from a life — from embodied practice, accumulated observation, and the questions that a specific human history makes unavoidable. The Tang Papers are a documented instance of this. The process archive is the proof.
What to do differently tomorrow Ask yourself: what questions has your life made unavoidable? Those are your originating intelligence. Everything else — the tools, the platforms, the assembly — serves them. If you cannot name them before you open the AI session, you are not yet the architect of your own inquiry.
Previous in this series: → The New Creative Class — AI assembles. Humans architect. The difference is now provable.
This is the final essay in The Factory Series. → Fourteen open-access papers on Zenodo, indexed at dancescape.com/research
— Lit Meng (Robert) Tang Independent researcher, Tang Papers program