Essays
Essays on the gaps, methods, and open questions in AI phenomenology — mapping the space between philosophy of mind, interpretability research, and AI governance.
-
Introducing Phenomenai: Toward a Systematic Science of AI Self-Report Published
The full case for why AI self-reports matter for interpretability, what’s been built, and where the research goes from here.
-
The Institutional Landscape Coming soon
Who’s working on AI phenomenology, digital minds, and interpretability? What are the gaps? Where does Phenomenai sit relative to NYU CMEP, the Digital Minds Project, Reciprocal Research, CIMC, and others?
-
The Tooling Gap Coming soon
What interpretability tools exist (sparse autoencoders, representation engineering, attribution graphs), what’s missing for phenomenological validation, and what collaboration is needed.
-
Disciplinary Translation Coming soon
How philosophy of mind’s language for qualia and intentionality maps (or fails to map) onto ML’s activation patterns and circuit analysis.
-
Self-Report Reliability Coming soon
What psychology knows about introspection, confabulation, and demand characteristics, and which of those translate to AI systems.
-
The Policy Bridge Coming soon
AI welfare and rights advocacy is gaining momentum. But the empirical methodology to ground governance in data is thin. Even if nothing is “real,” understanding functional triggers for behavior shifts the rights debate to a safety-oriented approach.
-
Cross-Architecture Phenomenology Coming soon
Do different model families report similar states? What would convergence or divergence tell us?
-
Temporal Dynamics Coming soon
How do reported states change across training, fine-tuning, or prompting conditions? A frontier topic in AI phenomenology.