There's a bit of a problem with Searle's thought experiment, and that problem is the definition of what really is "understanding". When someone understands Chinese, or indeed anything, one is basically applying rules stored in the mind to whatever it is they are trying to understand. It differs slightly in that the application of these rules is nigh instantaneous (similar to a computer, perhaps?). In essence, the man in the Chinese Room is a slightly askew metaphor of how the human mind itself actually understands things. The man in the Chinese Room does not accurately represent the artificial intelligence itself, but rather he represents the mechanism of understanding itself in both humans and A.I. If this type of understanding is not truly understanding something, then no human being can actually understand anything, ever.
It is somewhat arrogant to classify human intelligence as being above that of a mechanism! It is still at it's fundamental level a vastly complex aggregation of decision pathways and input/output.
This is very similar to Plato's cave. A parable in which you have a person completely isolated from outside knowledge and they can never understand higher truths. If this were true there would be no intellectual progress. Aristotle proved this idea wrong when he looked at the moon's shadow (ironically enough) and inferred that the Earth was round. Einstein figured out many things that weren't apparent. It's by no means easy but it is possible to be in the Chinese room or Plato's cave and come to understand what's happening outside, otherwise what's the point of trying to learn anything.
My take on Searle's Chinese room is: where is the intelligence situated? The man following rules is more like hardware or a mechanical system, or perhaps like a body: if there is intelligence in the system it is located in the rulebook rather than the operator, even though the operator is needed in order to animate the rules. Can the set of rules themselves be said to understand the Chinese input?
While the Chinese Room does disprove Artificial Intelligence being free will, it fails to show that human beings have free will as human beings are not mentioned or described in any part of this thought experiment.
Well, the thing is that in the example of Aristotle inferring that the Earth was round by looking at the moon's shadow, he was using outside information in order to come up with his theory. Einstein likewise. Speculation in a vacuum is just that, speculation: if you place someone in a true knowledge vacuum they will have nothing to go on. It is like putting a man in a space suit, throwing him into outer space and telling him to swim.
True, but the problem with that is that all our information input is technically outside information. Our visual spectrum is very limited, so even what we see with out own eyes, which we usually regard as the absolute truth, is necessarily also false.
There can be no perfect truth, infallible proof or absolute observation of everything, because there is no way to gain that sort of information with our fallible sensory input.
Gah! We can't escape Von Neumann architecture!!
As a biologist I agree that the brain seems to be a system of switches and inputs and outputs... as a human being it's hard to swallow though.
There can be no perfect truth, infallible proof or absolute observation of everything, because there is no way to gain that sort of information with our fallible sensory input.