Nina Miolane
Nina Miolane is the speaker in Claire Isabel Webb & Nina Miolane: The Geometry of Consciousness, where she presents a research program for a Mathematical Theory Of Intelligence. The source frames her work as a search for equations and geometric principles that can explain common structure across biological brains and artificial neural networks.
Research Position
Miolane’s key move is to shift attention from isolated neurons toward Population Coding and Neural Geometry. In the talk, spatial-navigation circuits and artificial networks trained on comparable tasks both show torus-shaped activity, making the Spatial Navigation Torus a test case for whether intelligent systems converge on efficient representational geometry.
She is careful about the boundary between intelligence and consciousness. The source presents Consciousness Measurement as an open problem where sleep-state geometry, replay, affect, and AI consciousness can be studied without claiming that a single metric decides whether a system is conscious.
The refreshed podcast export also makes her engineering implication clearer: if large networks discover useful geometry through training, smaller AI systems may need those geometric principles built into their architectures when data is limited.
Connections
- Claire Isabel Webb - interviewer for the Long Now conversation.
- Long Now - host context for the talk.
- Mathematical Theory Of Intelligence, Neural Geometry, and Population Coding - core research frame presented in the source.
- Spatial Navigation Torus and Fourier Spatial Encoding - main worked example and explanatory hypothesis.
- AI Interpretability By AI, Representation Learning, and World Models - adjacent AI themes the source extends through geometry and prediction.