We can't let the machines evolve and think for themselves, that's how they take over the world! In all seriousness though, the concept of creating new and better robot designs through an evolution like process is pretty ingenious. It is a little strange though because the actions of the robots are defined by predetermined responses based on stimuli. This is part of their software. This brings up a very interesting discussion about free will. I think that a common argument would be that humans have free will and that the robots don't; however, one could also argue that humans don't have free will either or that the robots have just as much free will as us (these could be the same thing but just stated differently depending on how you look at it).
The robots have essentially limited functions: executing a certain code, evaluating the efficacy of such code, and then writing new code based on this outcome. But what defines the concept of free will is essentially the complexity of the second step, which asks the pertinent question of: how can this code be best evaluated? One could, in theory, construct a matrix to determine the weight of each associated characteristic, but this matrix would be directly related to the first weight factors, the first input into the robot. This matrix becomes refined with each iteration, but once again these weights become directly dependent on the first choices. As such, the robot is still vitally recursive. This reaches back to the fractals: the behavior of an individual is both known and not, being dependent on the first configuration. Since no basis exists for this configuration, and with the programmer being ideally distant from the robot, one cannot be distanced from the concept of both free will and fate being intimately linked.
Self aware robotics is one of the scariest things to think about. The idea that we can create a robot that can think for itself and improve upon the work we've already done about it is really scary. Have you ever seen the Terminator? That's Skynet right there. Like Zach said, the actions of programs and responses of robots are predetermined, coded by man... But what happens if robots can do such a thing in the future? Will they be animalistic and try to do everything to maintain their survival and growth? This is much like the Matrix, and is the only way I see self aware robots trying to do anything dangerous.
Maz, Yes the robots actions at this time are predetermined by man (so we think), however I would challenge us to consider that perhaps this simulation is just a metaphor for man, in that all we are are self aware robots. If, as you say, we do not really have free will and are completely controlled by synapses in the brain, then we, much like your hypothesis about the robots, have self replicated in an animistic manner with the sole intention of doing everything in our power to maintain our survival and growth. So if we did it, why is it scary to think that robots may do the same? Seems like the natural order of things to me.
Oh my the intelligence of robots..... It really is a terrifying thing to see a robot that is able to learn and understand things. I can't help but think of the movie I-Robot. Computers today already seem to be able to do things that we can't. They can't "think" but they can "remember" more than we can, or at least we can retrieve far more information from a computer than we can from our own memories. Watching that one robot flop around was kind of funny for a while until I understood the significance of it. The robot figured out on its own how to move forward. It wasn't programmed to do so. Is it possible that one day robots could learn how to do more things than humans could? Could they become smarter than us? Take over? I'm not sure, but we definitely need to be careful!
I really do not understand how robots work or how these even work in the video, especially how are they have the able to move without a cue... Anyways it really is the best idea to try and create better robots by letting them evolve versus designing. Nature has proved itself to work for millions of years and the thought that us humans could create working robots of nature's likeness is almost ridiculous. I'm wary that this will work but I think that it is definitely a huge step in the right direction- mimicking nature in order to achieve stable, working products. I find this similar to the fractals and how the African tribes utilized a part of nature they knew was sustainable and used it in their own human structures.
I don't know if anyone has read the story "Do Androids Dream of Electric Sheep?", but it's the novel that the movie "Bladerunner" is based on. The stories in these focus on what it means to be human and what it means to be an android and pretend to be human. The whole concept of self-awareness in robotics reminds me of this. If we're moving in this direction of creating learning robots, won't we eventually be creating human analogs?
We can't let the machines evolve and think for themselves, that's how they take over the world! In all seriousness though, the concept of creating new and better robot designs through an evolution like process is pretty ingenious. It is a little strange though because the actions of the robots are defined by predetermined responses based on stimuli. This is part of their software. This brings up a very interesting discussion about free will. I think that a common argument would be that humans have free will and that the robots don't; however, one could also argue that humans don't have free will either or that the robots have just as much free will as us (these could be the same thing but just stated differently depending on how you look at it).
ReplyDeleteAh. the stuff of many long philosophical discussions….court battles, spiritual conversions!
DeleteThe robots have essentially limited functions: executing a certain code, evaluating the efficacy of such code, and then writing new code based on this outcome. But what defines the concept of free will is essentially the complexity of the second step, which asks the pertinent question of: how can this code be best evaluated? One could, in theory, construct a matrix to determine the weight of each associated characteristic, but this matrix would be directly related to the first weight factors, the first input into the robot. This matrix becomes refined with each iteration, but once again these weights become directly dependent on the first choices. As such, the robot is still vitally recursive. This reaches back to the fractals: the behavior of an individual is both known and not, being dependent on the first configuration. Since no basis exists for this configuration, and with the programmer being ideally distant from the robot, one cannot be distanced from the concept of both free will and fate being intimately linked.
ReplyDeleteSelf aware robotics is one of the scariest things to think about. The idea that we can create a robot that can think for itself and improve upon the work we've already done about it is really scary. Have you ever seen the Terminator? That's Skynet right there. Like Zach said, the actions of programs and responses of robots are predetermined, coded by man... But what happens if robots can do such a thing in the future? Will they be animalistic and try to do everything to maintain their survival and growth? This is much like the Matrix, and is the only way I see self aware robots trying to do anything dangerous.
ReplyDeleteMaz,
DeleteYes the robots actions at this time are predetermined by man (so we think), however I would challenge us to consider that perhaps this simulation is just a metaphor for man, in that all we are are self aware robots. If, as you say, we do not really have free will and are completely controlled by synapses in the brain, then we, much like your hypothesis about the robots, have self replicated in an animistic manner with the sole intention of doing everything in our power to maintain our survival and growth. So if we did it, why is it scary to think that robots may do the same? Seems like the natural order of things to me.
Oh my the intelligence of robots..... It really is a terrifying thing to see a robot that is able to learn and understand things. I can't help but think of the movie I-Robot. Computers today already seem to be able to do things that we can't. They can't "think" but they can "remember" more than we can, or at least we can retrieve far more information from a computer than we can from our own memories. Watching that one robot flop around was kind of funny for a while until I understood the significance of it. The robot figured out on its own how to move forward. It wasn't programmed to do so. Is it possible that one day robots could learn how to do more things than humans could? Could they become smarter than us? Take over? I'm not sure, but we definitely need to be careful!
ReplyDeleteI really do not understand how robots work or how these even work in the video, especially how are they have the able to move without a cue... Anyways it really is the best idea to try and create better robots by letting them evolve versus designing. Nature has proved itself to work for millions of years and the thought that us humans could create working robots of nature's likeness is almost ridiculous. I'm wary that this will work but I think that it is definitely a huge step in the right direction- mimicking nature in order to achieve stable, working products. I find this similar to the fractals and how the African tribes utilized a part of nature they knew was sustainable and used it in their own human structures.
ReplyDeleteI don't know if anyone has read the story "Do Androids Dream of Electric Sheep?", but it's the novel that the movie "Bladerunner" is based on. The stories in these focus on what it means to be human and what it means to be an android and pretend to be human. The whole concept of self-awareness in robotics reminds me of this. If we're moving in this direction of creating learning robots, won't we eventually be creating human analogs?
ReplyDelete