Foresight Institute Logo
Image of nano

Can “artificial life” evolve intelligence? An update

Artificial life from a digital sea (Image: Gusto Images/SPL)

An article in New Scientist with the optimistic title “Artificial life forms evolve basic intelligence” gives an update on how two specific examples of computational artificial life is doing in terms of evolving to have more interesting behavior.  An excerpt:

Brains that have been evolved with HyperNEAT have millions of connections, yet still perform a task well, and that number could be pushed higher yet,” he says. “This is a sea change for the field. Being able to evolve functional brains at this scale allows us to begin pushing the capabilities of artificial neural networks up, and opens up a path to evolving artificial brains that rival their natural counterparts.

See the comments after the article for useful discussion.  A field to keep an eye on.  —Chris Peterson

4 Responses to “Can “artificial life” evolve intelligence? An update”

  1. Hephaestus Says:

    It seems to me this HyperNEAT might be usable for the recognition portion of a mind machine interface. If the evolutionary algorithms were set up to evolve for speed and accuracy of pattern matches it might be possible with an Nvidia GPU and little hardware for input to get this done in short order.

  2. John Burch Says:

    Hi, I realize a brain is a neural net but it seems to be organized with higher level connections which means, to me, that you will not get higher intelligence out of a single bowl of neural stew. Maybe exceedingly good pattern recognizers, or a balance control system, but not a person. In fact I would not want to create an entire brain(mind) as one neural net. It would be as much a black box as any of these neural nets already walking around (people). I’d rather see a mind put together out of subsystems and then you might have some hope of knowing what was going on inside and might give you a handle on how to affect it’s operation. For more perspectives and a suggestion for how to create self-awareness in a computer, visit Nanofuture2030

  3. Marktron Says:

    Ok, the artificial intelligence in fact is the organization of any kind of knowledge (it could be the action or reaction, information etc..) to solve a task nonplanned satisfactorily, it means that an automatic machine is not intelligent, it´s just an automation, The intelligence requires to jump the database of posible things or programmed things.
    Yeah an artificial cell must be intelligent in a evolution process or adaptation to something.
    The organic cell is not intelligente with some virus like VIH.
    Long life to the articial life!!!

  4. Delphi Says:

    what if these new “brain structures,” or wetware or whatever one wishes to call them could become “vessels” for dimensional entities?

    one can’t engineer coherent thought, so how could an assembler “build” one?

    i mean, what is driving a machine’s path to sentient thought?

    i’m just saying…

Leave a Reply