>I have heard this claim frequently, that intelligence is "embodied" in a way that computers overlook, but if that turns out to be critical, well, who is to say that something like this "embodied" context can't also be modeled computationally?
Well, Searle argued against it when presenting the case for the Chinese Room argument, but I disagree with their take.
I personally believe in the virtual mind argument with an internal simulated experience that is then acted upon externally.
Moreso, if this is the key to human like intelligence and learning in the real world, I do believe that AI would very quickly pass by our limitations. Humans are not only embodied, but we are prisoners to our embodiment and we only get one. I don't see any particular reason why a model would be trapped to one body, when they could 'hivemind' or control a massive number of bodies/sensors to sense and interact with the environment. The end product would be an experience far different from what a human experiences and would likely be a super organism in itself.
Well, Searle argued against it when presenting the case for the Chinese Room argument, but I disagree with their take.
I personally believe in the virtual mind argument with an internal simulated experience that is then acted upon externally.
Moreso, if this is the key to human like intelligence and learning in the real world, I do believe that AI would very quickly pass by our limitations. Humans are not only embodied, but we are prisoners to our embodiment and we only get one. I don't see any particular reason why a model would be trapped to one body, when they could 'hivemind' or control a massive number of bodies/sensors to sense and interact with the environment. The end product would be an experience far different from what a human experiences and would likely be a super organism in itself.