"There’s a scary lesson in these contests between the grandmaster and his soulless opponents. We are sharing our world with another species, one that gets smarter and more independent every year. Though some people scoff at the idea that machines could become autonomous, remember it wasn’t long ago that almost no one thought a computer would ever beat a human chess champion. Could we ever face anything akin to the horrendous sci-fi nightmares that we see in “Terminator 3”? In the long run, it’s well worth worrying about. But the machines aren’t worried at all."

Machine vs. Man: Checkmate

by Alex Roland and Philip Shiman. MIT Press, June 2002, ISBN 0-262-18226-2. 440 pages.

This is the story of an extraordinary effort by the U.S. Department of Defense to hasten the advent of "machines that think." From 1983 to 1993, the Defense Advanced Research Projects Agency (DARPA) spent an extra $1 billion on computer research aimed at achieving artificial intelligence. The Strategic Computing Initiative (SCI) was conceived as an integrated plan to promote computer chip design and manufacture, computer architecture, and artificial intelligence software. What distinguished SCI from other large-scale technology programs was that it self-consciously set out to advance an entire research front. The SCI succeeded in fostering significant technological successes, even though it never achieved machine intelligence. The goal provided a powerful organizing principle for a suite of related research programs, but it did not solve the problem of coordinating these programs. In retrospect, it is hard to see how it could have.

In Strategic Computing, Alex Roland and Philip Shiman uncover the roles played in the SCI by technology, individuals, and social and political forces. They explore DARPA culture, especially the information processing culture within the agency, and they evaluate the SCI?s accomplishments and set them in the context of overall computer development during this period. Their book is an important contribution to our understanding of the complex sources of contemporary computing.

Strategic Computing: DARPA and the Quest for Machine Intelligence, 1983-1993.

... At that point, as far as I'm concerned, they're hired.

Inside PARC

Pierre Moulin has a paper entitled "Information-Theoretic Analysis of Information Hiding" in the March 2003 issue of IEEE Transactions on Information Theory. You can find this paper online here, along with his other related work.

Excerpts from the abstract of the paper: An information-theoretic analysis of information hiding is presented, forming the theoretical basis for design of information-hiding systems. ... The host data set is intentionally corrupted, but in a covert way, designed to be imperceptible to a casual analysis. We formalize and evaluate the hiding capacity ... [which] is the value of a game between the information hider and the attacker. It is shown that many existing information-hiding systems in the literature operate far below capacity.

Information-Theoretic Analysis of Information Hiding

You'd think that this was late-breaking stuff, but the original paper is three years old.

ScienceDaily News Release: Cracking Data Hiding: Theory Can Help Disable Terrorists' Messages