David Thorstad asks if bounded agents can have both accuracy and coherence

Analogue computation, properly understood, is the best candidate for characterizing neural computation, argues Corey J Maley

Mario Hubert on how an investigation of the initial-value problem challenges our view of electromagnetism