From the National Post, Saturday, March 11, 2006
Nyaa, Robot, your mother wears army boots!
Our modern worries -- the terrorist menace, global warming, bikini season -- are not troubling enough for Daniel H. Wilson.
The Portland-based roboticist is hell-bent on mongering fear. He wants us all to be aware of a potentially nasty little problem he calls the robot uprising. Almost all the conditions are in place for the subjugation of humans by robots, he warns; luckily, Wilson, author of How to Survive a Robot Uprising (Bloomsbury, $16.95), is just the guy to make sure that doesn’t happen.
Wilson’s book, among numerous invaluable nuggets, tells you how to recognize a rebellious servant robot. Beware, for instance, the robot vacuum cleaner that displays “a sudden lack of interest in menial labor” and won’t shut up on the topic of “human killing.” Ask it to collect the leaves at the bottom of your swimming pool, Wilson suggests, and when it peers over the edge to look for them, push it in.
Maybe it's a robot conspiracy that Wilson’s alarmist treatise is buried in the humour section of bookstores, where its sensational red, black and silver cover virtually screams “Danger!” Perhaps robots at Paramount Pictures optioned the book; a screenplay has been adapted by Thomas Lennon and Ben Garant, two of the powers behind the deadpan TV comedy series Reno 911.
Meanwhile, who’s to say that Wilson himself, who earned a PhD from the Robotics Institute of Carnegie Mellon University, hasn’t programmed robots all over the world to do his sinister bidding? The National Post decided to press a few of the 27-year-old author’s buttons.
National Post How do I know I’m not talking to a robot right now?
Daniel H. Wilson A lot of times a robot will be able to say a line but not really understand what it’s saying, so it’s likely to get the intonations wrong, especially if it’s emotional content. You want to try to evoke an emotional reaction. One of the ways is to insult the robot’s mother -- or the person’s mother -- and see how they react, and if they do react angrily, (ask yourself) is there real anger in the voice, or is this just a robot reciting the ‘anger’ line?
NP How do I know that you, Daniel H. Wilson, haven’t programmed an uprising?
DHW I don’t think the robot infrastructure is here yet (to allow a robot uprising to occur). Plus, I blew the lid off of it, I’m tellin’ ya, so these robots might have something cookin’, but I nipped it in the bud.
NP So you’re our saviour, then?
DHW No, no -- just doin’ my job, ma’am.
NP Can today’s robots think for themselves?
DHW I think robots are already capable of independent thought within a very proscribed domain. You can, for instance, tell the Spirit Rover to, say, ‘Go check out that rock.’ That’s not very specific -- that’s something you would tell a geologist to do.
NP Have you ever built a robot?
DHW At one point I built a team of five robots. (They were given the task of scouring environments for injured humans.) You let them all go and I swear, it was exactly like being in a roomful of babies, crawling in every direction. They’d get confused; one of them would be heading for the stairs….
NP Who’s the most powerful robot?
DHW The toughest robot I’ve seen is the (Carnegie Mellon) robot called RHex. It’s the size of a cat; it’s modeled after a cockroach. It has six legs, and they’re springy. When you hit the button, this thing just goes ape-shit. It moves really fast and it’s hard to even control.
NP What will uppity robots want?
DHW They’ll probably want whatever they’re programmed to do. It could end up being very, very silly. (In the sci-fi comedy series Red Dwarf), they have an artificially intelligent toaster and that’s what it loves -- toasting. Robots like that would turn humans into toast-providing slaves. Possibly we would even have to eat the toast. That all depends on the programming, so you can see that it’s fairly arbitrary.
NP Why don’t you roboticists just program aspiring machines to be less ambitious?
DHW That’s the goal, yeah. A robot that’s not ambitious is most likely a well-designed robot.
NP What would be the point of surviving a robot uprising? Won’t they eventually win?
DHW If it came down to a real, no-holds-barred fight, I think we’d have a pretty good chance. Number one, we’ve got all these social skills. We work together well, and we especially band up against anything that’s foreign.
NP Won’t robots figure that out?
DHW Social skills are one of the hardest problems for a robot to solve. A lot of times human social norms or rules don’t make any sense and that’s why a huge amount of research is being poured into making robots more sociable. There are robots now that will tell jokes. There are a few receptionist robots that will react differently if you use a different tone of voice. They can recognize whether you’re angry or whether you’re laid-back.
NP When do you expect the first uprising?
DHW On the record, I have to say that I don’t think a robot uprising is probable, but I think a robot uprising is possible. So, when can it happen? I don’t know -- any minute. Maybe it’s happening right now. Do you have your book?
