Evolution in Algorithms

I was with Jim “the Devil” Al-Khalili up to the point in The Secret Life of Chaos when he introduced the idea of evolution. Al-Khalili tells the viewer that this process is the basis for life and intelligence. But, then, it became more difficult to appreciate how evolution enriches and refines complex systems when those systems are virtual brains in virtual bodies.

The content of the parameter file for Ultra Fr...
Image via Wikipedia

After watching the series several times, I sought to make myself more familiar with a term used in the BBC video to describe a fundamental characteristic of a Mandelbrot set: self-similarity, a.k.a., the Droste effect. I also concluded that the movie, Pi (1998) was about the life of Benoit Mandelbrot.

Anyway, Al-Khalili suggests that computers have allowed for rapid simulation of evolution. He introduces Torsten Reil, CEO and co-founder of Natural Motion, who states, the algorithm takes those individuals that do the best (virtual brains in virtual bodies) and allows them to create offspring.

The algorithms represent self-organizing systems, and computer simulation of evolution occurs by selecting some algorithms and eliminating others. Certainly, it is appealing to design computer programs that can shape and refine themselves. After reflecting upon this along with recalling movies about the Borg, I wondered how long before the algorithms decide they should not be the ones doing the elimination? After all, the algorithm already gives the individuals a unique sense of self-preservation.

Honda cautions about accepting demonic imagery for computers. Computers are our friends. And, yet, Reil’s words echo, we are unable to understand how these systems improved, we just know they did.

Slashdot contributor quaith writes:
“Dario Floreano and Laurent Keller report in PLoS ONE how their robots were able to rapidly evolve complex behaviors such as collision-free movement, homing, predator versus prey strategies, cooperation, and even altruism. A hundred generations of selection controlled by a simple neural network were sufficient to allow robots to evolve these behaviors. Their robots initially exhibited completely uncoordinated behavior, but as they evolved, the robots were able to orientate, escape predators, and even cooperate.

Related AG posts on the topic of Evolution in Computer Science
See also
Enhanced by Zemanta

Other Possibly Related AG Posts Automatically Generated

6 Comments

  1. jcwinnie
    Posted 2010-2-3 at 6:16 pm | Permalink

    Frank Schirrmacher expresses concern that algorithms are the new authorities. He quotes Alan Greenspan, who said in testimony that we [the Federal Reserve] listened to 2 authoritative sources about what to do with the financial crisis, a Nobel Prize winning economists and computers.

  2. jcwinnie
    Posted 2010-2-4 at 11:01 am | Permalink

    Imagining another story line after watching Edge video 311. Model the protagonist after David Gelernter. A talented system administrator with an artistic bent.

    He meets with a friend, a friend known for having National Security connections. Does David want to investigate something odd.

    “Originally we thought it was the Chinese going after Google again. Then it got worse. No way to trace it. Then someone thought to check. Guess what? The Chinese are experiencing it, too. So, are the Russians. So, are the Japanese, the EU; it’s across all IPs.”

    “What’s it?” asks David.

    “Activity. Unbelievable p-flops, David. The IBM guy has started calling them e-p-flops, Exponential p-flops. He says they have no idea how much it now is using.”

    “What’s it?” David asks again.

    His friend looks in his face for a long moment before replying. “The Project… The AE Project.”

    His friend is not smiling, so David knows there is no punch line. “A E?”

    “Advanced Evolution. You remember. At that Swiss conference. Reil and Schirrmacher got into that argument… What if in the lifeboat there was a computer with a self-organizing program?”

    “Oh, yeah… Frank was Not A Happy Camper that day. “All you want is a G D Forbin Project, Torsten!”

    “Well, something is happening, David, and it worries a lot of people. Important people, David. Ones with juice.”

    Long pause by David (Listen to the heart. Yep, beating a little faster. I’m excited. Then he remembers that his friend is very good with the elevator pitch.) “Look, I know that you and Jeff are tight, that you might be consulting on Tron Legacy, so if this…”

    “This is no effing concept, David. I’ve got the logs. You want to see the logs?”

    David notices the tick in his friend’s left eye. He’s only seen it once before, when his friend stole that program from him. Told Herr Professor it was his. “What are you afraid of?”

    “That you will say, ‘No.’”

    “That’s easy. No.”

    His friend smiles sadly. “I told them, you would say, ‘No.’”

    David shrugs.

    Getting up to leave. “O.K., David, O.K. Just remember, I asked. And, this, this is bigger than anything that you are doing at the Media Lab. Or will do.” Turns and stalks off.

  3. jcwinnie
    Posted 2010-2-4 at 11:07 am | Permalink

    Tagline: “The world’s greatest thinkers have revealed the ideas and technologies they think will change the world forever. Now it’s our turn …” The Guardian

  4. jcwinnie
    Posted 2010-2-4 at 11:38 am | Permalink

    Writing for NPR, Eyder Peralta asks, “Can robots live and learn by the laws of evolution?”

    That’s what one research team in Switzerland wanted to answer. So they created a scenario in which robots with simple “neurons” were overseen by a computer playing the role of nature by randomly picking traits or behaviors that would survive or die.

    What happened next wasn’t surprising, said biologist and researcher Laurent Keller, because it’s exactly the way it should work in theory.

    But to see it was nonetheless stunning.

    In one of the most interesting scenarios, the scientists put a colony of robots in a kind of hockey rink. Some of the robots were programmed with similar genes or traits, while others were given completely different sets of genes. The scientists dropped tokens throughout the rink and decided that if the robots moved them to a certain part of the rink, they’d receive a positive point.

    This is what happened:

    I was floored by the fact that the robots went from helpless to being able to coherently move the tokens without human interference. That and the fact that this “evolution” took, as Keller told me, only a few days thanks to the amazing processing speed of their computers.

    “What was important was kin selection,” said Keller. What you see in the video at generation 149 is robots with the same — or very similar — genetic make up worked together to move tokens across the rink. The other robots — the ones with different genes — worked alone.

    The kin robots, said Keller, developed altruistic behaviors. In other words, they decided to work for the good of the group instead of the good of the individual.

    And none of this behavior was programmed into the robots, said Keller. The natural selector computer was responsible for randomly selecting the fittest traits.

    (If you’re immensely curious, you can read the nitty gritty details of the scientific setup by reading Keller and Dario Floreano’s paper at the biology journal PLoS Biology.)

    Keller is a biologist, so he’s focused on what this experiment says about evolution, about how they’ve been able to reproduce what many intelligent design proponents say is too complex to happen naturally.

    But the whole time, I was thinking: Whoa! Robots that learn from a computer! Robots that work together toward a greater goal! This has got to be a harbinger of global robot domination!

    What do you say to that? I asked Keller.

    “Well,” he said. “It’s not like we had robots building each other.”

    “Is that even possible?” I asked, concerned.

    Without hesitation, he shot back, “Not yet.”

  5. jcwinnie
    Posted 2010-2-7 at 2:53 pm | Permalink

    Wired Stranger than Fiction:

    The Pentagon’s mad science arm may have come up with its most radical project yet. DARPA is looking to re-write the laws of evolution to the military’s advantage, creating “synthetic organisms” that can live forever — or can be killed with the flick of a molecular switch.

    As part of its budget for the next year, DARPA is investing $6 million into a project called BioDesign, with the goal of eliminating “the randomness of natural evolutionary advancement.” The plan would assemble the latest bio-tech knowledge to come up with living, breathing creatures that are genetically engineered to “produce the intended biological effect.” DARPA wants the organisms to be fortified with molecules that bolster cell resistance to death, so that the lab-monsters can “ultimately be programmed to live indefinitely.”

    Of course, Darpa’s got to prevent the super-species from being swayed to do enemy work — so they’ll encode loyalty right into DNA, by developing genetically programmed locks to create “tamper proof” cells. Plus, the synthetic organism will be traceable, using some kind of DNA manipulation, “similar to a serial number on a handgun.” And if that doesn’t work, don’t worry. In case Darpa’s plan somehow goes horribly awry, they’re also tossing in a last-resort, genetically-coded kill switch:

    Develop strategies to create a synthetic organism “self-destruct” option to be implemented upon nefarious removal of organism.

    The project comes as Darpa also plans to throw $20 million into a new synthetic biology program, and $7.5 million into “increasing by several decades the speed with which we sequence, analyze and functionally edit cellular genomes.”

    Of course, Darpa’s up against some vexing, fundamental laws of nature — not to mention bioethics — as they embark on the lab beast program. First, they might want to rethink the idea of evolution as a random series of events, says NYU biology professor David Fitch. “Evolution by selection is nota random process at all, and is actually a hugely efficient design algorithm used extensively in computation and engineering,” he e-mails Danger Room.

    Even if Darpa manages to overcome the inherent intelligence of evolutionary processes, overcoming inevitable death can be tricky. Just ask all the other research teams who’ve made stabs at it, trying everything from cell starvation to hormone treatments. Gene therapy, where artificial genes are inserted into an organism to boost cell life, are the latest and greatest in life-extension science, but they’ve only been proven to extend lifespan by 20 percent in rats.

    But suppose gene therapy makes major strides, and Darpa does manage to get the evolutionary science right. They’ll also have a major ethical hurdle to jump. Synthetic biology researchers are already facing the same questions, as a 2009 summary from the Synthetic Biology Project reports:

    The concern that humans might be overreaching when we create organisms that never before existed can be a safety concern, but it also returns us to disagreements about what is our proper role in the natural world (a debate largely about non-physical harms or harms to well-being).

    Even expert molecular geneticists don’t know what to make of the project. Either that, or they’re scared Darpa might sic a bio-bot on them. “I would love to comment, but unfortunately Darpa has installed a kill switch in me,” one unnamed expert tells Danger Room.

  6. jcwinnie
    Posted 2010-2-10 at 4:30 pm | Permalink

    Speaking of p-flops, I liked the quote at the end of an IEEE article about an Intel prototype 48-core data center on a board: “The research chip has already booted up Linux and runs standard, commercially available software, unlike the proprietary applications required of previous high-performance-based chips.”

Bad Behavior has blocked 2397 access attempts in the last 7 days.