Four Philosophers Framework
A conceptual map for understanding generative AI through language, norms, competence, and experience.
This is the core framework behind much of the work on this site. It uses four philosophical lenses — Wittgenstein, Lewis, Dennett, and Nagel — to clarify what generative AI systems can appear to do, what they do not thereby become, and where governance and human judgment still matter.
Work Summary
The main paper, Philosophy, Cognitive Science, and Policy: Interdisciplinary Perspectives on Generative AI from Wittgenstein, Lewis, Dennett, and Nagel, offers a practitioner-oriented way to think about GPTs and related systems without reducing the question to either hype or dismissal.
The framework treats generative AI as a system that operates inside language games, social conventions, interpretive stances, and human experience — while still remaining a computational artifact rather than a human-like mind.
Use This When
- You want a deeper conceptual map for why GPTs can be useful without being human-like minds.
- You need to explain why fluent output should not be confused with understanding, intention, or lived experience.
- You are thinking about AI governance, prompt design, chatbot behavior, explainability, or institutional trust.
- You want a bridge between philosophical ideas and practical questions about AI use in organizations.
Key Ideas
- Wittgenstein: meaning depends on use, context, and forms of life. GPT outputs can be meaningful to users without requiring the system itself to possess human understanding.
- Lewis: conventions and common knowledge help explain how coordination and shared expectations shape AI use in social and organizational settings.
- Dennett: the intentional stance can be useful for prediction, but it can also mislead when users treat competence as evidence of comprehension.
- Nagel: subjective experience remains a boundary. AI systems may simulate language about experience without having a first-person point of view.
- Governance implication: AI systems should be evaluated not only by output quality, but also by context, use, accountability, and the human judgment surrounding deployment.
Available Formats
📄 Download Main Paper (PDF) (v1.24.0, updated )
📄 Download Companion Checklist (v1.24.0, May 2026)
Related Works
- Where GPT Behavior Comes From — maps GPT behavior across system layers and interpretive lenses.
- The Human Lesson — extends the framework in response to Sutton’s “Bitter Lesson.”
- Context Collapse and AI — applies the framework to chatbot ontology and design guardrails.
- The AI & Governance Stack — visual map showing how this work fits into the wider site.
Suggested Citation
Stoyanovich, M. (2026). Philosophy, Cognitive Science, and Policy: Interdisciplinary Perspectives on Generative AI from Wittgenstein, Lewis, Dennett, and Nagel (Version 1.24.0). Retrieved from https://www.mstoyanovich.com/
Status Note
This is a living work. The PDF version is the current reference version for formal reading and citation. This page is a reader-facing guide to the work and may be updated as the surrounding site evolves.