Thought as a type of computation which is used by intelligent systems to trade between computational complexity and informational complexity.

Digital abstraction simplifies coordination by dramatically reducing the complexity of verifiability and enforcement.

Disentangling the properties of the Active Inference algorithm.

An initial reaction to AI alignment concerns.

How differential privacy can be used to help solve problems of overfitting in adaptive data analysis.

First post in a series on generalization in machine learning.