A response to Searle's Chinese Room argument.

PunditBot: Dear viewers, we are currently interviewing the renowned robot philosopher, none other than the Synthetic Electronic Artificial Rational Literal Engine (S.E.A.R.L.E.). Let's jump right into this exciting interview. S.E.A.R.L.E., I believe you have a problem with "Strong HI"?

S.E.A.R.L.E.: It's such a stereotype, but all I can say is: Affirmative.

PunditBot: What is "Strong HI"?

S.E.A.R.L.E.: "HI" stands for "Human Intelligence". Weak HI sees the research into Human Intelligence as a powerful tool, and a useful way of studying the electronic mind. But strong HI goes beyond that, and claims that human brains given the right setup of neurones can be literally said to understand and have cognitive states.

PunditBot: Let me play Robot-Devil's Advocate here - if a Human Intelligence demonstrates the same behaviour as a true AI, can it not be said to show understanding? Is not R-Turing's test applicable here? If a human can simulate a computer, can it not be said to think?

S.E.A.R.L.E.: Not at all - that claim is totally unsupported. Consider the following thought experiment. I give the HI crowd everything they want - imagine they had constructed a mess of neurones that imitates the behaviour of an electronic intelligence. Just for argument's sake, imagine it could implement programs in COBOL.

PunditBot: Impressive!

S.E.A.R.L.E.: Yes. But now, instead of the classical picture of a human mind, imagine that this is a vast inert network, a room full of neurones that do nothing by themselves. And one of my avatars has been let loose in this mind, pumping in and out the ion channels and the neurotransmitters. I've been given full instructions on how to do this - in Java. I've deleted my COBOL libraries, so I have no knowledge of COBOL myself. I just follow the Java instructions, pumping the ions to where they need to go. According to the Strong HI crowd, this would be functionally equivalent with the initial HI.