This interview with Hubert Dreyfus (just the parts about computers: part 1, part 2. via Continental Philosophy) briefly outlines one of the major criticisms leveled against artificial intelligence: computers will never be intelligent because our intelligence is based upon our physical interactions in and with the world. Very briefly, our intelligence is fundamentally tied to our bodies because it is only through our bodies do we have any interaction with the world. If we separate our intelligence from the body, as in the case with computers, then whatever it is that the computer has, it is not intelligence because intelligence only refers to how to bodily interact with the world.
As Dreyfus says this problem is attributed to a Merleau-Ponty extension of Heidegger and the only proposed solution is to embody computers by providing them with a full representation of world and body. I don’t think there is generally much faith in this solution; I certainly don’t have much faith in it.
However, this bodily criticism is a straw man. Computers have ‘bodies,’ they are definitely physical things in the world. But what of the physical interactions required for intelligence? Computers interact with the world: computers are affected by heat, moisture, dirt, vibration, etcetera. The only differences are the actual interactions that computers have as compared to humans: we experience humidity one way and they experience it differently. So yes, computers will have different interactions and hence they will never have the same intelligence that we have, but that does not imply that computers cannot have an embodied intelligence. It only means that computer embodied intelligence will be significantly different than our own intelligence. Therefore the above argument against computer intelligence only applies to those people who are trying to replicate perfect human intelligence and does nothing against people trying to create intelligence in computers.
For example, light-skinned and dark-skinned people have very slightly different physiologies. Now I see the above argument as saying that someone of different skin color cannot have the same sort of intelligence that you have because their interactions with the world are inherently different. Sure, everyone experiences things slightly differently due to having different bodies, but to claim that this creates incompatible intelligences is obviously wrong: No one on the face of the earth would be able to communicate with each other due to everyone being physically unique. Computers may be physically different to a greater extent, but this does not impact intelligence.
The criticism of computer intelligence based upon the need for a body is no more than subtle techno-racism.
