Revisionist William Gibson

When I made yesterday’s entry, I intended to talk a bit about William Gibson‘s appearance at the Boulder Book Store last week. The reading took place in the large upstairs room of the store, which Gibson referred to as a “ballroom,” and the crowd had standing room only. I was fortunate enough to get there early since I work in Boulder, and I was rewarded with not only a good seat, but a low number for the book signing line.

Gibson read a chapter from his latest book, Pattern Recognition. Well, initially, I thought he was going to cough a chapter of it; something got hold of his throat and the water he drank to soothe it seemed to make the problem worse. Despite this, Gibson croaked on, and eventually he seemed to find his voice again. It was a bit surreal for a moment, more than a hundred of us sitting there silently listening to his amplified coughs reverberating through the store. I expected people to involuntarily clear their throats, but if anyone did, I didn’t hear it. There wasn’t much we could do, but it was an awkward moment in which I wanted to do something to help him, and didn’t.

The chapter was short, a bit shorter than the Q & A session afterward. I did manage to pose a question to him. The exchange went something like this (paraphrased, since I didn’t take notes):

SJ: You seem to have a talent for putting words together in unusual ways.  Are there any exercises you do to cultivate that talent?

WG:  Revision.  It’s all revision.  I write the drafts, then rework it until it’s right.  

That makes sense. Although I’m a big believer in revision, for some reason I continue to have the notion that Great Writers spew complete paragraphs forth, Zeus-like, from their heads. This personal myth is probably fueled by reports of Isaac Asimov and Ray Bradbury writing without revision, and surely, the more one writes, the easier it becomes to pre-form great sentences. But that’s no reason to expect that all successful writers write without revision; I’m glad Gibson reminded me that it’s not the case in his life. That makes me feel a little better about the level of his writing, too. It’s a bit more attainable now.

I haven’t read all of Pattern Recognition yet, but what I did read grabbed me. It’s Gibson’s first novel set in the present day, and it’s interesting how it still feels very much like a Gibson novel, despite the fact that no one has an input jack in his head and there are no major characters that are AIs. The evil international mega corporation backdrop is still there, but this time it takes the form of corporate branding á la Tommy Hilfiger. Heavy fragment use in writing, much like this sentence.

One notable difference is that September 11, 2001 plays a direct role in the development of the main character, Cayce Pollard. At his appearance, Gibson spoke at length about how he had completed 100 manuscript pages of the book before the attacks, and how the real world events utterly destroyed those pages of fiction. To be believable, Cayce Pollard had to be completely re-imagined as a result of the attacks, and the first 100 pages had to be re-written. He did so, and I think the revision contributed to the dark feel of the book. Reading this, I get the feeling it’s not such a big jump from Pattern Recognition to the shatterglass worlds of Neuromancer or Virtual Light.

I don’t know if that’s a good thing or not.

The Tomorrow Makers

The Tomorrow Makers

By: Grant Fjermedal

Type: Non-fiction survey with touches of biography

Setting: N/A

Description:

Based on interviews with some of the most notable cognitive science researchers in the country and their students, Fjermedal’s book walks the line between non-fiction and biography. Without going into the messy details, he shows that many people in the world believe in the possibility (probability?) of building robots and computers “smart” enough to hold carbon copies of a human mind and continue its thinking processes after the download. 

Comments:

Fjermedal realizes something that not many other survey interviewers do: in an institutional setting, the big names aren’t necessarily the ones who do the most work. Fjermedal not only concentrates on the big fish in the AI pond, such as Marvin Minsky, Joseph Weizenbaum, John McCarthy, Allen Newell, Gerald Jay Sussman and Danny Hillis. He also focuses on the students. Many of them stay up for days at a time working on projects with the kind of dedication that most people don’t give to their careers. They deserve a round of applause, and Fjermedal gives it to them. The student viewpoint is also interestingly fresh because they are accomplished dreamers. They are not afraid to speak of what they think will happen twenty or thirty years down the road. Whereas some university professors will pad their opinions and say, “Well, that might happen someday,” the students respond with, “That will happen. And I’ll do it.” This approach may not be entirely realistic, but reality is not necessarily a good culture for new ideas. 

Recommendations:

This book, combined with Machinery of the Mind, by George Johnson, works well as a non-technical survey of the directions of artificial intelligence and the people driving there. Fjermedal goes a little more into the personalities and the distance possibilities than Johnson does, but the two books give a consistent view of the field. Specifically, Fjermedal tries to show why the researchers are trying to create intelligent computers and shows the energy with which they are working.

Machinery of the Mind

Machinery of the Mind

By: George Johnson

Type: Non-fiction survey of AI

Setting: N/A

Description:

During a year of intensive study, George Johnson travelled around the country to conventions, interviewed prominent researchers in the field of cognitive science and read just about everything there is to read on the use of machines to model human intelligence. 

Comments:

Although a few years out of date, Johnson’s book is still a fresh, easy to understand look at the advances in the new science of artificial intelligence (or cognitive science, as some researchers prefer to call it.) Johnson is a good writer, and is obviously an intelligent man. He understands the concepts presented in his book, even though the knowledge comes from many different fields, all of which meet at the center of cognitive science. MotM makes a valient effort to present different sides of the artificial intelligence issue, devoting time to the “engineers,” the people concerned with presenting a working product that doesn’t necessarily have to model human intelligence, and the “scientists,” the heavy-hitters in the AI world who are trying to accurately model the workings of the human mind, whether it is practical or not. I got the impression that Johnson favors the pure research side more than the commercial aspects. 

Recommendations:

This is a great introductory text to artificial intelligence research. I wish I had had it when I started reading Gödel, Escher, Bach many years ago!.

The NeXT Step (Vorticism 1991)

The NeXT Step?
(Vorticism 1991)

Thousands of programmers
Typing in unison
Hacking in UNIX on
Sleek black machines.

Weizenbaum warns us that
AI’s not where its at
Threats to humanity
Garnish the seams

Of these C algorithms
And LISP subroutines.
Icons to silicon
Bury our dreams.

Massively parallel
Spells analog’s death knell;
Hypercubes emulate
Turing machines

In a giant array
of AI DNA,
Launching electrons like
Silical genes.

Cognitive science has
Taught us to ask ourselves:
When does Intelligence
Make its own means?

Sufficiently complex
Parallel neural nets
Attain self-awareness at
Some point, it seems;

What right as Creators
Have we to berate or
Suppress the inception of
Silicon Dreams?

— Stace Johnson, 1991