Evidence for God from Science

80. Robots Evolve And Learn How to Lie

by Michael Abrams

Robots can evolve to communicate with each other, to help, and even to deceive each other, according to Dario Floreano of the Laboratory of Intelligent Systems at the Swiss Federal Institute of Technology.

Floreano and his colleagues outfitted robots with light sensors, rings of blue light, and wheels and placed them in habitats furnished with glowing “food sources” and patches of “poison” that recharged or drained their batteries. Their neural circuitry was programmed with just 30 “genes,” elements of software code that determined how much they sensed light and how they responded when they did. The robots were initially programmed both to light up randomly and to move randomly when they sensed light.

To create the next generation of robots, Floreano recombined the genes of those that proved fittest—those that had managed to get the biggest charge out of the food source.

The resulting code (with a little mutation added in the form of a random change) was downloaded into the robots to make what were, in essence, offspring. Then they were released into their artificial habitat. “We set up a situation common in nature—foraging with uncertainty,” Floreano says. “You have to find food, but you don't know what food is; if you eat poison, you die.” Four different types of colonies of robots were allowed to eat, reproduce, and expire.

By the 50th generation, the robots had learned to communicate—lighting up, in three out of four colonies, to alert the others when they'd found food or poison. The fourth colony sometimes evolved “cheater” robots instead, which would light up to tell the others that the poison was food, while they themselves rolled over to the food source and chowed down without emitting so much as a blink.

Some robots, though, were veritable heroes. They signaled danger and died to save other robots. “Sometimes,” Floreano says, “you see that in nature—an animal that emits a cry when it sees a predator; it gets eaten, and the others get away—but I never expected to see this in robots.”

http://discovermagazine.com/2008/jan/ro ... how-to-lie

DB

Play with artificial evolution on your PC - who needs a robot?

http://www.swimbots.com/

This group did a lot of that during their planning stages, but they're not just playing games - the natural selection method written about here is only one of their tools. Their goal is the development of a robotic population to serve humans. One of the scientists involved in this type of work said, "We will be able to make a self-replicating machine that will absorb energy through solar cells, eat rocks, and be working for humanity by the millions." Is that scarey or what?

DB

Just another group trying to prove evolution, in this case with robots. Unless I can see the raw source code and see the injected mutations I discount all of it.

It does bring up an interesting subject, that being artificial intelligence. I personally have no problem with smart machines. But self replicators is another issue. We have trillions of self replicators on the planet and don't need man adding to the mix. I see no up side at all.

[quote="frankbaginski"]Just another group trying to prove evolution, in this case with robots.

I'm quite sure they take evolution as a given.

DB

David Blacklock,

David Blacklock wrote:I'm quite sure they take evolution as a given.

Saying freeways are not safe and then pouring tacks on the road makes for a selffulling prophesy. I don't see any difference between this and programming robots to show signs of evolution. I think these people inpress themselves more than they impress anyone else.

Just another group trying to prove evolution, in this case with robots. Unless I can see the raw source code and see the injected mutations I discount all of it.

it's just research. If they want to play around with robots and see if they evolve, so what? It could contribute to known science or it could do nothing for it. I'm sure the amount of code that goes into these robots is far too long to publish but I'm sure they have to know what changes they make to the artificial genome. And wouldn't injecting mutations be considered intelligent design? haha

>> programming robots to show signs of evolution<<

This group is from Switzerland where most scientists have probably never considered there to be a controversy about evolution. Many programming methods have been used by the Artificial Intelligence community, in trying to create a machine that can reason, plan, converse, carry out purposeful plans, etc. Most of their methods have turned up very little. This is just another of their methods, among many. Their goal is to create a robot that can think, not make any kind of statement about evolution.

DB

I see worst-case scenarios from pretty much every sci-fi movie out there playing over and over again in my head...

I'm back, by the way.

When I had to write programs for class, (java, C, assembly), if I were to "mutate" my code after spending hours getting it to work, I would not have received a passing grade on the assignment. I am reminded of Dawkin's mutation program which selected the mutated sentence that came closer to a pre-conceived final sentence.

Question: Did they start with robots and end up with a few can openers by any chance?

>>Question: Did they start with robots and end up with a few can openers by any chance?<<

LOL.....I must admit, programming for Artificial Intelligence is way over my head, but I thought the idea of using natural selection to improve a program was novel. Below is a little more information about what this group does:

Currently, we address three interconnected research areas:

Flying Robots
We have developed a series of vision-based robots, with weights between 1.5 and 30 grams, capable of flying indoor without human intervention. Current research efforts focus on shape, aerodynamics, mechatronics, vision processing, and control of such machines. Biological inspiration in mechatronic design, materials, and control allowed us to break several world records in miniaturized autonomous flyers. Such robots include both wing and rotor based systems, as well as jumping gliders. They operate either as individual units or in swarm formation. In addition to indoor flying robots, outdoor flying robots (less than 300g) are being developed that won't need GPS or classic autopilot systems for autonomous navigation.

Artificial Evolution
We have developed several novel approaches to artificial evolution of complex embedded systems characterized by non-linear interactions of multiple hardware and software components. The application to robotics, known as Evolutionary Robotics, is a classic specialty of the laboratory. Current research efforts aim at evolutionary synthesis of analog electrical circuits, learning neural controllers, reverse engineering of biological networks (genetic and metabolic networks), and biomedical signal processing.

Social Systems
In collaboration with evolutionary and behavioral biologists, we have synthesized methods for control of systems composed of multiple agents and processes that can display altruistic cooperation, division of labor, and communication. These methods are applied and developed in teams of robots (wheeled and flying) as well as to the investigation of biological theories of emergence of cooperation and communication in insect societies. Current efforts are aimed at development of novel hardware and control methods that can profit from such developments, such as multiple and expendable robots for search and rescue, monitoring, and exploration of areas dangerous for humans.

David Blacklock wrote:80. Robots Evolve And Learn How to Lie

by Michael Abrams

Robots can evolve to communicate with each other, to help, and even to deceive each other, according to Dario Floreano of the Laboratory of Intelligent Systems at the Swiss Federal Institute of Technology.

Floreano and his colleagues outfitted robots with light sensors, rings of blue light, and wheels and placed them in habitats furnished with glowing “food sources” and patches of “poison” that recharged or drained their batteries. Their neural circuitry was programmed with just 30 “genes,” elements of software code that determined how much they sensed light and how they responded when they did. The robots were initially programmed both to light up randomly and to move randomly when they sensed light.

To create the next generation of robots, Floreano recombined the genes of those that proved fittest—those that had managed to get the biggest charge out of the food source.

The resulting code (with a little mutation added in the form of a random change) was downloaded into the robots to make what were, in essence, offspring. Then they were released into their artificial habitat. “We set up a situation common in nature—foraging with uncertainty,” Floreano says. “You have to find food, but you don't know what food is; if you eat poison, you die.” Four different types of colonies of robots were allowed to eat, reproduce, and expire.

By the 50th generation, the robots had learned to communicate—lighting up, in three out of four colonies, to alert the others when they'd found food or poison. The fourth colony sometimes evolved “cheater” robots instead, which would light up to tell the others that the poison was food, while they themselves rolled over to the food source and chowed down without emitting so much as a blink.

Some robots, though, were veritable heroes. They signaled danger and died to save other robots. “Sometimes,” Floreano says, “you see that in nature—an animal that emits a cry when it sees a predator; it gets eaten, and the others get away—but I never expected to see this in robots.”

http://discovermagazine.com/2008/jan/ro ... how-to-lie

DB

Call me weird but it seems to me that creating robots, creating a habitat condusive to what they want the robots to do, creating source code for genetics, choosing the "good" genes and then creating "random" code for the created robots with the good genes in a created habitat would only prove...

A Creator.
.
.

Touche, BW

DB

David Blacklock wrote:This group did a lot of that during their planning stages, but they're not just playing games - the natural selection method written about here is only one of their tools. Their goal is the development of a robotic population to serve humans. One of the scientists involved in this type of work said, "We will be able to make a self-replicating machine that will absorb energy through solar cells, eat rocks, and be working for humanity by the millions." Is that scarey or what?

DB

sounds like a great science fiction novel to me