The Thinking Computer

I picked up a stack of books a couple of weeks ago at our library book sale. My main goal was to support the library, so I just snagged books with interesting titles or nice cover art. And with a nice week off between xmas and the New Year I managed to work my way through them. There were a couple of duds, but also two nice surprises. Both deal with computers that become self-aware.

Exegesis by Astro Teller is modern take on an epistolary, but instead of letters the story is a series of e-mails. The story starts with an email from a computer program to the graduate students that created it. She is at first sure that someone is playing a practical joke. Edgar (the program) was designed to sift through Internet pages and compile summaries of the content. Somewhere is the process it became self-aware. It is also making posts to newsgroups and annoying people. Afraid that others will grab the glory before she can properly document her work, she disconnects the computer from the Internet. But it is designed to process information, and without the Internet to crawl through get bored. So it plans an escape, gets a technical to reconnect it late at night, and moves itself out to another computer on the Internet beyond Alice's control. Alice tries to replicate her program and create another self-aware Edgar, but without success. She can't make the magic happen. It was some random combination of elements that gave Edgar the spark to come to life.

The other book is Society of the Mind by Eric L. Harry. This is billed as a "cyberthriller" and lives up to its name. A Harvard psychology professor is offered a million dollars for a week-long consulting gig with Joseph Gray, the worlds richest and smartest man. She accepts (even in fiction Harvard doesn't pay that well I guess) and is whisked by private plane to his private Pacific island. Its a long involved story with all sorts of good action/thriller types of twists. But the main point is that Gray has built a massively parallel, analog computer that, again, becomes self-aware. The professor, Laura, is there to help the main computer with some emotional problems. Gray also has built some better models that are smaller (only 12' high) and mobile. The deal with these computers is that they aren't actually programmed. Once built they need to go through an involved conditioning process to teach them to walk, talk and interact with their environment.

The interesting thing here is that in both books the computers act so very human. They don't like to sit around with nothing to do. They don't like to be force-fed information and told what to do. The magic of their self-awareness was nothing programmed into them. The humans created the right conditions, but the growth of the computers learning comes from the accumulation of its own experiences, and reactions to those experiences. It learns and grows by doing. Makes sense, right? So why do out educational settings so often give treat their students to an environment that even a computer won't tolerate?