Born in 1956
. His father, Bill Hillis was an Air Force epidemiologist, and the family moved frequently on the trail of hepatitis
His mother taught him mathematics and his father taught him biology... a near perfect setup for his future life. As a child, was extremely interested in Boolean algebra (based on and, or, and not operations).
Hillis went to MIT in 1974 and planned to major in neurophysiology (he really wanted to understand how the brain worked). He created computers and programs for children that didn't know how to read or write, and was allowed in the AI lab.
His understanding of neurophysiology (although limited) and his mother and father teaching him biology and mathematics set him up perfectly to be a pioneer in AI. He thought of a brain that was essentially decentralized (like the human brain... it has different parts working for different subgoals... while each subgoal contributes to the main goal... like partitioning of a brain).
Hillis wanted to get thousands (originally millions!) of processors and hard drives, hook them up in a massive net and create AI. It would be called the Connection Machine. He envisioned a computer emulating a human brain but with much more processing power, and near infinte memory (and near perfect too!).
This is called massive parallelism. It is controversial because of Amdhal's law. It basically says:
"This doesn't work because although you have the processing power, getting the data that needs to be processed to all the different processors takes too long, and so does sending the data back"
In 1983, he founded Thinking Machines in Cambridge, Massachusetts. The president of CBS was a major investors, and DARPA bought the first machine. Richard Feynman worked for the company to help it with the short term memmory in network data.
Hillis worked heavly in neural networks (a set of programs that behave as neurons... talking to each other, doing calculations and recieving stimuli).
For more information on this sort of subject, I recommend you check the soft links, and research... Bart Kosko - Chief thinker of fuzzy logic
Fuzzy Thinking - Instead of if and than... if and than!
Neural Networks - Programs designed to emmulate neurons
Beowulf Clusters - Distributing processing power
AI - Artificial Intelligence
Douglas B. Lenat - Pioneer of AI
And for fun Neuromancer - Cyber punk novel about AI