As a followup to James' post on parallel chemical detection via films on CD surfaces:

Slashdot | Beer-Coated CDs are Optical Biocomputers: "A DJ and scientist in Melbourne whose research is in the area of communication through biological cells, serendipitously created an 'optical biocomputer' when he spilled beer on his CDs and left them over night. The resulting fungus that formed distorted the sound of the CDs in interesting and meaningful ways. Here's some of his research, and some media samples which include mp3s of the distorted music."

This guy seems to be flinging his work a little far. He's written five papers on this for four different conferences in July. I haven't read them yet, but it looks like they are wildly speculative. From what I can see, he's grown some stuff on some CDs and DVDs and verified that they screwed up the data, but he doesn't indicate (on the web site at least) whether he's actually done any useful computation. Maybe there's more in the papers...