It’s one thing to imagine these things are going to be developed and quite another to read about them. Wired carried an article yesterday titled “Experimental AI Powers Robot Army” (Link). If you’ve not read it, you should. If you need convincing, try these excerpts:
The software is a type of neural network with two special features. One introduces perturbations, or “noise,” into the network so that existing ideas get jumbled into new forms. The second is a filter that assesses the new ideas against existing knowledge and discards those that are unsuitable.
This reminds me of Rudy Rucker’s book “Software“.
Self-learning and adaptability will be the key to success, and this is where the Creativity Machine excels. Give it any set of robotic limbs and it will master locomotion within minutes without any programming, swiftly finding the most efficient way of moving toward a goal.
More importantly, it’ll never test positive for steroids.
Perhaps the most impressive — and spookiest — aspect of the project is the swarming behavior of the robots. In computer simulations, they acted together to tackle obstacles and grouped together into defensive formations where needed, Thaler said. They also worked out how to deal with defenders, and spontaneously devised the most efficient strategy for mapping their environment, he added.
(The comic relief has already been expended – it’s time to get worried here.)
Thaler declined to describe his results in detail, but said his system has produced unspecified “humanlike capabilities.”
“I can relate the results of virtual-reality simulations, where swarms of Creativity Machine-based robots have deliberatively sacrificed one of their kind to distract a human guard, enabling the remainder to infiltrate a mock facility,” he said.
Oh great. And I bet they don’t stay in the same place when you try to play the level over either, do they?
Thaler’s current project, which should be completed over the next six months, will develop a piece of software called CSMARRT (for Creative, Self-Learning, Multi-Sensory, Adaptive, Reconfigurable, Robotics Toolbox). The software can be used to design and model virtual robots that can be placed in virtual environments to learn and develop.
Just don’t give them access to an RP machine.
One, by the Defense Threat Reduction Agency, which is tasked with dealing with WMD threats, commences in fiscal year 2007 for “technology necessary for robotic systems to attack tunnel complexes” and another for “microdamage technologies” for “very small robotic weapons.”
And this is where I say, “I told you so” (cuz somewhere I wrote about wasp-like weapons, but I forget where).
This seems an opportune time to talk about weaponised killer bees.
Swarm-behaviour AIs have been around for a while now though. You don’t want to believe any company that’s angling for a Homeland Security grant.
Cute. Reminds me of those… *ahem* … homing penis things.
It’s not the swarming behavior that’s so interesting to me; I’ve written about swarming UCAV’s. It’s not the collaborative work either, because I’ve posted videos here of polybots doing stuff like riding a tricycle. It’s those things and on top of the evolutionary stuff that’s so fascinating.
btw, feel free to list the features – or future feature – of those killer bees. I’m curious.
Oh, indeed, it’s fascinating stuff; I’m just a little concerned about the gap between simulated robots and actual, instantiated robots. Though NASA have certainly had success with their evolutionary simulators.
My bees, to be quite honest, are not terribly smart; they will certainly chase you down and sting you to death if you annoy them by getting too close, but they won’t, say, build a hive if you disappear, send out bees to see where you are, determine the best method of killing or annoying you by evolutionary algorithms and communicate what you look like to other hives by means of dancing. I fear that the Lindens might not be terribly keen on bees which not only reproduce but hold a grudge.
(I did consider using evolutionary algorithms for my automatic vehicles, though, to try to work out the best method of avoiding them falling foul of sim crossings, since there seems to be no rhyme or reason to the matter.)
“I’m just a little concerned about the gap…”
You and me both, although I’d use the adjective “decreasing” in front of “gap”.
As for the bees, I’ve entertained the thought of making a virtual cruise missile that hunts down avatars and upon finding them delivers a payload of insults. That shouldn’t be too bad and within the ToS I’d think (assuming it didn’t proceed indefinitely – it could last for a minute then de-rezz… at which point another missile is launched from home, of course).