Secure, reliable (Re: Distributed computing)

A couple of months back, someone (Adam Back, perhaps?) made mention of work on secure, reliable distributed computing on untrusted networks of untrusted machines. I seem to remember mention of sending out bogus computations in order to obscure the actual computation being performed. I was never able to follow the reference, since all the info available was the name of a Japanese professor, with no mention of the university or of any publications I could look up. Does this work ring a bell with anyone else? nathan