Thursday, December 15, 2005

I got a job, Engineering Scientist Associate, at UT. I'd probably be remiss to post the exact department on my blog. If you are one of the three people who reads this blog, and you want to know, please e-mail me and I'll tell you. Actually, two people already know, so Greg, if you want to know, e-mail me.

I am excited about this. I was worried I'd get a job that would force me to forget most of what I studied: physics, math, digital signal processing. This job will give me opportunities to maintain and build upon my academic foundation. My three year, unrequited love affair with the wave equation can continue; the job involves acoustics. On Tuesday I spent my coop bookstore rebate on a book called "fundamentals of physical acoustics." First chapter, the wave equation. It's easy enough to see that functions with an x-ct (position minus speedconstant times time) parameter shift right on the x axis as you increase t. A pulse glides right on the x axis like a wave, or a lightbike from "Tron." The wave equation is a partial, second order differential equation that, when you plug in tron-bike equations and differentiate like a good calc II student, works. Admittedly, I am still baffled by how mathematicians derived the wave equation. When I really get it--and I must--I will post a long blog entry or pdf document to explain. It has been too long in coming.

More to the point people, CLOSURE. In my Morita pepper blog entry I hinted that I feel that my life is almost too complicated to explain, but here are the lowlights, in sequential order: demeaning office jobs, lost my father to cancer, worked for a three-person multimedia start-up and the boss committed suicide, i made a serious go of it as a tech writer and got laid off, i went back to school and worked my way up from the lowest-level EE courses, selling my condo in the process. A decade of suspense. Of course, I am leaving many things out, many good things, and I am not so naive to assume that everything will be better here on out, but I think at some point you can safely say that you paid some dues.

Currently reading a fascinating book, Jeff Hawkins' "On Intelligence." He invented the Palm Pilot and went on to study neuroscience. His book attempts to explain the human mind as if it were an operating system. 50 years ago, a lot of his analogies and suppositions would have seemed trivial. Such as, intelligence is being able to infer from a siren that a fire truck is coming, we don't have to actually ID the truck visually. Big deal. But the point he makes is that our ability to predict outcomes is distinctively human, possible due to our large neocortex, the "new rind" around our more primitive inner brain. The upshot of computers and AI might not be that we eventually write intelligent computer programs, but that by writing computer programs we become able to appreciate the aspects of intelligence that are not "computational." Hawkins and his pals are out to build neocortex-like systems. I think this has fascinating implications for the field I am getting into. I would bet money that the kinds of systems I will be working on rely on instantaneous detection and recognition of patterns in the environment, and that the idea of recognizing patterns by correlating with stored "memories" of prior patterns has not been tried. I hope I'll be able to blog about it.