Of What Do the Sheep Dream?

I've got a big bowl full of ever-so-slightly expired cottage cheese (smells fine, tastes weird), a good solid hour of free time, and not a single interesting accomplishment in the last month. Blog potential: very high.

The next three weeks are a time of elevated tensions in the Thornton house. Years ago, when Julie started her current job, she assumed the role of Open House organizer person. She performs this duty admirably - handling lots of tricky logistical and bureaucratic nuisances that make it possible for the students to present a vision of the college and themselves that is both energetic and learn-tastic.

Unfortunately, by the time she gets home she tends to be a bit low on mojo. I've been trying to medicate that with cake, but now she's starting to think she's eaten too much cake. I'll need to find something stronger.

Eventually, the cascade of pre- open house fun always ends in our garage. She tends to give access to the students for the purpose of assembling their parade float. We usually even buy them pizza. Multiple times! (two nights) Currently, I have the second half of a shelving project lingering in most of the garage. The pressure to finish up before they get here is probably a good thing. I'm more worried about the squat rack.

It turns out the squat rack is big. Really big. I'm going to have to stuff/slide/push it somewhere. The worst part about that is that then all those students will think I don't use my squat rack. Ungh. Students...

Moving on. A couple weeks ago, I had what I believe was a B-variant of influenza. It wiped me out for a few days and in the end it took a week and a half to get over, but I did manage to get something useful out of it. I read my first book on Julie's iPad. A paid-for book even. I felt very fancy.

Book Review

The book in question was Do Androids Dream of Electric Sheep? Most people (including myself until a couple weeks ago) know it vaguely as the inspiration for the 1982 Harrison Ford film Blade Runner. A movie held in renown for it's ability to evoke a seemingly impenetrable sense of boredom right up until the last two minutes when all of the subtle concepts smack you in the face and leave you saying, "Whoa, dude, that was almost worth it."

Most people don't make it that far and if you skip ahead it doesn't happen. "They" keep releasing very different cuts (5 by my count) and the same thing happens every time. It's pretty amazing.

However, the book is not the movie. In the case of the book, the first 95% was really interesting and the end was not very enjoyable. The concept we find ourselves exploring is the same concept you explore when you join a new forum or try to post a link to Facebook. In essence, are you a machine?

Machines absolutely hate strike-through text.

The book was written in 1968. Like most authors operating during the dawn of computer science, he thought we'd be a bit further along in the artificial intelligence by now. Humanoid robots are common and they are physically indistinguishable from humans (aside from a bone marrow test). Mentally, they lack empathy. Unfortunately, they occasionally go nuts, start killing people, and try to return to Earth. (The book is set in 2021 and we've already colonized two planets.)

Our protagonist, Rick Deckard, is a robot hunter. He finds these unruly computers and puts them out of commission - with very realistic explosions of blood and gore. His hardest job is to figure out if he's about to retire (i.e. blast in the face with a laser cannon) a sneaky robot, or a human. Fortunately, he is armed with a sequence of questions he can use to test empathy based on the subject's responses. His struggles administering the tests and beyond - robots can have false memories, how do I know if I'm a human?!?! - give us a lot to think about.

Sidebar: why does it matter? Seriously. When I got into the evacuation modeling business, I started telling a "joke" that goes like this:

some guy: Does your simulator model human behavior?
me: It models movement and includes enough AI to keep them from bouncing into stuff...much.
same guy: But does it correctly handle the situation where [something elaborate].
me: Uh, probably not. Do you have any thoughts on how it should handle that?
some guy: It should use a neural network and fuzzy logic and work just like the human mind, and...
me: Well, if we did that and we were successful we could never turn it off.
some guy: Eh, why not?
me: It would be murder.

(Haha- wait wtf?)

So you see, I have a problem with the book's premise. Here we've got these creatures that go rogue and escape their tyrant owners just so they can try to live out the remaining portion of their meager four-year lives in relative freedom. When Deckard redecorates the inside of his patrol car with blasted android brains, we're expected to take comfort in the fact that it was "just a robot." Meh. End Sidebar.

So the majority of the book is a thriller with some very exciting moments where a character completely pivots and invokes a major thought paradigm shift. In the very end, I think were supposed to see one of those on Deckard when he supposedly "gets religion" - based on the immersive Hollywood empathy experience that addicts most of the human population - but in reality he probably just freaks out and starts having hallucinations.

Final analysis: it was neat for a computer person. It was also neat for someone who enjoyed the last two minutes of Blade Runner and wanted to know what they were trying to do. If you aren't already into that kind of thing, I don't think I'd recommend it. My favorite part of the book was the forward by Roger Zelazny. He played a creative writing game called "Thirteen ways a garbage disposal repair man" and it was awesome. I'll have to try it sometime.

Comments