Panta Rhei[1]
Practical essays, maps, and tools for making generative AI understandable, governable, and useful.
Who This Site Is For
This site is for executives, technologists, governance professionals, consultants, educators, and serious readers trying to make sense of generative AI in real organizations. The work here combines practitioner experience, philosophical interpretation, and governance discipline to ask a practical question: how can we use powerful AI tools without confusing fluency for understanding, automation for judgment, or interface behavior for institutional trust?
The AI & Governance Stack
A skeleton key to this site’s corpus: a seven-layer diagnostic map for locating where AI problems tend to live — in meaning, system behavior, interface design, organizational reality, work practice, governance handoff, or human judgment.
📄 Download Stack Map (PDF) (v3, May 2026)
Start Here
- New to this perspective? Start with the Four Philosophers Framework™ to explore how Wittgenstein, Lewis, Dennett, and Nagel help us understand AI.
- Want a practical tool? Download the CONTEXT Prompting Framework for a tested and structured way to get better results from AI.
- Prefer interaction? Try the chatbots: Four Philosophers Chatbot or CONTEXT Chatbot.
Ephemera
Exploratory papers to stretch your thinking. These pieces bring classic philosophy, cognitive science, and AI governance into practical conversation.
Suggested path: Philosophy, Cognitive Science, and Policy → Where GPT Behavior Comes From → Old Tools, New Eyes → The Human Lesson → The Question Concerning Learning.
Optional branch: Context Collapse.
-
Philosophy, Cognitive Science, and Policy Start here
A Four-Philosophers Framework for Understanding Generative AI
This conceptual lens draws on Wittgenstein, Lewis, Dennett, and Nagel to examine language, reasoning, intention, and consciousness in AI — a compact lens for meaning, norms, competence, and experience.
📄 Download Paper (PDF) (v1.23.9, updated )
“So What? The 10 Things to Implement” — a one-page practical checklist. -
Where GPT Behavior Comes From Map
Four Philosophers, a linguistic stack, and a practical map of deployed systems
When a GPT “hallucinates,” where in the stack did things go wrong? This essay links a four-philosophers lens (Wittgenstein, Lewis, Dennett, and Nagel) to a linguistics stack, and then maps both onto a plain-text GPT system. The map separates the system into three interdependent lanes: Control, Core, and Outside-core.
📄 Download Paper (PDF) (v1.3.1, updated )
-
Old Tools, New Eyes Persistence
Edgerton’s The Shock of the Old through Wittgenstein, Lewis, Dennett, and Nagel.
Most technology talk fixates on invention; most of our lives are governed by what persists. This essay reads David Edgerton’s The Shock of the Old through the Four Philosophers Framework™ to explain why meaning, convention, cognitive bias, and layered experience keep old infrastructures in charge even in the age of AI. It offers a lens for seeing AI deployments as layers on top of legacy systems rather than clean breaks from the past.
📄 Download Paper (PDF) (v1.4.2, published )
-
The Human Lesson Counterweight
A response to Sutton’s “Bitter Lesson” through Wittgenstein, Lewis, Dennett, and Nagel.
Why competence without comprehension strains explainability, responsibility, and civil trust.
📄 Download Paper (PDF) (v1.6.2, updated )
-
Context Collapse and AI Guardrails
Ontological collapse, LLMs, and the Four Philosophers Framework.
Examines how treating chatbots “as if” they were interlocutors collapses the distinction between tools and persons, using Wittgenstein, Lewis, Dennett, and Nagel to diagnose this ontological collapse and to propose design and governance guardrails.
📄 Download Paper (PDF) (v1.4.2, updated )
-
The Question Concerning Learning Deep dive
Babich, Heidegger, and the Enframing of Intelligence.
Prior question: what is learning? Babich/Heidegger on enframing and its implications for modern AI.
📄 Download Paper (PDF) (v1.0.1, updated )
Briefs & Whitepapers
For leaders and strategists. These memos help organizations align on AI roles, risks, and institutional readiness.
-
The Employment Game
Work as a socially embedded practice in the age of AI
A conceptual brief arguing that jobs are not merely bundles of tasks, but socially embedded practices carried out within a wider “employment game” of permissions, exceptions, coordination, accountability, and practical know-how. Introduces a five-layer model and applies it to benefits administration and cybersecurity / GRC.
📄 Download Brief (PDF) (v1.8.1, published )
-
A Constitution for a Chatbot
Vendor constitutions, interface ontology, and local controls
An analytic governance brief arguing that vendor “constitutions” can reduce behavioral risk while amplifying ontological risk—unless translated into local, auditable, role-bound practice. Includes a reusable local AI constitution template, an illustrative regulated instantiation, and a diagram of the full “constitution stack.”
📄 Download Brief (PDF) (v1.1.3, published )
-
The Anders Addendum
When Action-Power Exceeds Imagination
A governance-oriented addendum introducing Günther Anders’s “Promethean gap” as an escalation trigger: when system scale, speed, or blast radius outstrips responsible human imagination, ownership, and answerability.
📄 Download Brief (PDF) (v1.0.1, published )
-
AI Adoption is Mostly “The Shock of the Old”
A quick crosswalk between my “Old Tools, New Eyes” paper and Every’s field report on what’s actually working in companies.
A short diagnostic memo mapping real-world AI adoption patterns to four philosophical lenses and to the “old layers” that usually determine success: data, identity/access, process legibility, and incentives/routines.
📄 Download Brief (PDF) (v1.0.1, published )
-
Leading Through Alignment
A Framework for the Chief AI Officer
Designed for executives, board members, and tech strategists to define and institutionalize the CAIO role.
📄 Download Brief (PDF) (v1.2.1, published )
Tools
Turn insight into action. These frameworks are designed to help you prompt, teach, or govern AI more clearly and intentionally.
CONTEXT Chatbot Prompt Framework
A Structured Prompting Model Rooted in Philosophy and Practice
The CONTEXT Prompt Framework supports clarity, ethical alignment, and iterative refinement in human-AI interaction.
Grounded in best practices from design, education, governance, and communication, it helps users structure prompts for
high-quality, purpose-aligned AI outputs.
Unlike ad hoc prompting, CONTEXT has been tested in a small comparative benchmark: in a
recent benchmark
recent benchmark comparing CONTEXT to structured prompting heuristics from OpenAI and Anthropic (HHH), CONTEXT performed strongly across three real-world tasks — UX design, education, and risk governance. This is not a universal validation claim, but it suggests CONTEXT is a practical, testable structure for prompt design.
Whether you’re building chatbots, teaching AI literacy, or managing compliance-sensitive AI use, CONTEXT offers a practical,
tested structure for prompt design that reduces iteration and increases output quality.
📄 Download CONTEXT Guide (PDF)
Automate the Repeatable, Own the Judgment
A three-layer model for using AI without outsourcing responsibility
A compact tool for designing AI-enabled workflows: what to automate, what must remain human judgment, and how to govern the handoff so drafts don’t silently become “truth.” Includes a practical diagnostic checklist and governance posture.
📄 Download Tool (PDF) (v1.0.2, published )
AI Use Discipline Kit
A practical “how to use AI” discipline — aligned to the DOL AI Literacy Framework
A compact toolkit for using GenAI without outsourcing judgment. Includes a one-page discipline card, a copy/paste prompt scaffold, and a quick evaluation rubric for checking outputs before they become “truth.”
- Discipline card: a repeatable “stance → prompt → boundary → stakes” cycle.
- Prompt scaffold: a structured template for constraints, inputs, and success criteria.
- Evaluation rubric: a quick scoring check for accuracy, logic, fit, and risk.
📄 Download Toolkit (PDF) (v1.0.0, published )
Custom Instructions for GPT Assistants
Four-Philosophers Overlay and Governance Modes
A platform-agnostic instruction framework for conversational AI, adaptable to Microsoft Copilot, ChatGPT custom GPTs, and similar systems. The framework combines explicit knowledge-base validation, falsifiability checks, interpretive guardrails, and reasoning quality controls to improve transparency, trust, and decision support in complex or high-stakes use cases.
Chatbots: Four Philosophers Chatbot · CONTEXT Chatbot
About
This site is a personal exploration of ideas. I am not writing as an academic philosopher or machine-learning researcher, but as a practitioner using philosophy, systems thinking, and governance experience to make generative AI more intelligible and accountable in real settings. If you want to know more about my professional role, find me on LinkedIn.
Have a unique use case or experience with AI? I’d love to hear about it. Feel free to reach out and share for future inclusion.