Chalmers worked with Hofstadter while getting his PhD at Indiana. I know they've at least written one article together. It's in Hofstadter's book Fluid Concepts and Creative Analogies. The article is a criticism of traditional efforts in AI (GOFAI - Good Old Fashioned Artificial Intelligence, as John Haugeland calls it) to model high-level cognitive processes without also incorporating simple perception processes as well. Their claim is that both kinds of processes are required to accurately model human-style intelligence.
Also, I noticed you mentioned Andy Clark. I'm currently working on my MA Thesis which draws heavily from his work on the Classical / Connectionist debate (Microcognition being my primary text for this).