Computation - readings

What stuck out the most from the readings was the contradiction between computation's universality and limitations. If a Turing machine is capable of computing anything a current computer can, and computation is implementation-independent, cognition may theoretically be represented computationally. But this promptly brings up the symbol grounding issue: if computing is merely syntactic manipulation, when can we talk about semantics? How do we talk about it? Can pure computation ever provide meaning, or does cognition require something non-computational?

Turing's proof of uncomputable issues raises another question: are our cognitive boundaries the same as those of universal machines, or can humans access resources beyond computation? And Horswill's reminder that each century defines the mind in terms of its dominant technology raises the question of whether "mind as computation" is really a modern metaphor.

The larger issue seems to be simulation. A computational model can provide the correct results, even precisely mirroring behavior, but this does not imply that it is thinking. A simulation of cognition, like a simulation of fire, may fail to produce cognition. The concern is that by focusing solely on functional equivalency, we combine performance and process; mistaking the shadow for the item itself. If that's the case, the fundamental question isn't whether machines can simulate thinking, but whether simulation alone can ever pass the threshold into true cognition.