Let's talk about good ideas for a minute. Good ideas are like building blocks. Pretty much everything worthwhile, from a fun weekend to a breakthrough in medicine, begins when someone puts a good idea out there into the ether. Sometimes that idea stands alone, a self-sustaining entity. Sometimes it is met in kind with other good ideas, stacking upon one another, an endless bucket of Lincoln Logs.
For the most part, good ideas are instantly recognizable:
Hey, I've got an idea, let's get a puppy. — Good. Hey, I've got an idea, let's try putting some cheese on that burger. — Good. Hey, I've got an idea, let's wash our hands before we operate on people. — Good. Hey, I've got an idea, let's play ping-pong. — Good. Excellent, actually. Hey, I've got an idea, let's watch birds and then write down what birds we see. — Good? I guess? Hey, I've got an idea, let's have sex and eat candy. — Depends on several variables, but probably good. Hey, I've got an idea, let's build a swarm of robots. — Good. Wait, hang on, bad. Yeah! And we can design the robots to re-program themselves, and assemble more robots on their own. — Whoah! Stop right there, definitely bad. Then we can launch them throughout the solar system. — What th- Are you fucking kidding me? Mother of mercy, bad! Very, very bad!
Alas, that particular sequence of ideas is on its way towards reality.
The basic thought process at work here is not entirely crazy. Say you build a complex robot that can do a number of tasks. If it gets damaged, you're shit out of luck, especially if you just spent three years and hundreds of millions of dollars flying the thing to another planet. However, if you build a swarm of wee little robots, when one gets damaged it's no big deal.
The thing is, every time you need to update their programming, you'd have to go one by one through the whole swarm, and no matter how many interns you have at your disposal, that's going to eat up some serious time. (Plus, the interns will screw it up. You know they will.) So instead, imagine if the robots could re-program themselves to adapt to the situation. Encounter a problem, solve it, and move on. To further minimize the need for
human intern labor, the robots would be able to replenish their number by building more of each other.
Which is fine.
Except that it involves, wait for it ...
... building a swarm of problem-solving, self-propagating robots.
At some point, shouldn't certain word combinations send red flags cascading through the air? I really feel like I shouldn't have to be pointing this out. As soon as "swarm" and "robots" appeared in tandem, someone in the room should have suggested that they all take a walk around the block — stretch the ol' legs, maybe get a smoothie or something.
Look, I understand that there are marvelous applications for such a (shudder) swarm. The example given in the article is to send them into a collapsed building, where they can divide themselves into groups to search for survivors and check for further dangers.
It's just that, as project leader Klaus-Peter Zauner puts it, "with swarm robots, even if a percentage of them fails, they'll still be able to achieve their goal." Leaving aside for a moment the fact that I'm pretty sure someone named "Dr. Zauner" fought Green Lantern for several issues in the mid-70s, he basically just told us that the swarm will be unstoppable. As in, unable to be stopped.
The pièce de résistance in this Fellini-show of logic is that, naturally, swarm robots would be a great way to explore Mars. A lot of Mars missions have petered out because the rover went glitchy, but with a swarm? No problem.
Okay. You know how sometimes parents will send a kid off to college, and the kid comes back for Thanksgiving spouting off about their new world-view and complaining about the oppressive colonial tradition silently perpetuated by roasting a turkey, which oh by the way is a barbarous practice? Meanwhile the parents just want to smack the kid upside the head and eat pie?
Well think of Mars as college, only when the robot swarm comes home with some funny ideas about who's in charge, they won't have a head to smack and we won't be able to shut them off.
And there will be no pie.
Anyway, check out the video embedded in the article. I have to admit, when the little robots team up to drag away the bigger robot, it really is pretty cool.
It'll be a bummer when they figure out how to do the same thing to me.