Keith sat on the park bench and played guitar for coins. Sometimes I would take a break from reading microprocessor manuals and listen. Keith had paranoid schizophrenia. He could explain how the world worked: "There is a great international conspiracy..." he would say. Electromagnetic fields, government satellites, resonant dinner plates, you name it: he had it all figured out. This was back in the days of the 80386, when the CPU had only four levels of indirection in its addressing architecture. But something about the way he explained his world caused it to stick with me all this time.
Keith couldn't trust the conflicting information coming from different parts of the brain. He knew that he was vulnerable, and he clearly spent a lot of energy thinking about it. Does this not also describe our current relationship with data security? Our architectures have become so complex that they are inherently susceptible to internal schism, leaving us vulnerable to sudden manipulation by shadowy external forces. How does Keith's literal world view look today? Could a paranoid schizophrenia-based model of data security hold significant predictive power, grant new insights, or suggest a beneficial course of action?
Marsh has been desoldering computer circuits since he was a small child. He has built some stuff. He has broken some stuff. He has helped to fix some stuff. He spoke about it at Shmoocon last year. Speak, Marsh, speak! If my talk is accepted, I promise to get a real bio.