In this chapter he talks about Searle’s famous "Chinese room" thought experiment.

http://en.wikipedia.org/wiki/Chinese_room

I always thought that a good (at least partial) response was one that Blackburn gives, which is that’s like saying that neurons don’t understand Chinese. I assume Searle anticipated this response, I haven’t read the literature.

Blackburn mentioned this, but using Wikipedia’s description:



This doesn’t seem all that convincing to me, and Blackburn finds this to be unsatisfactory.

I certainly see a difference between saying a person “understands” Chinese while a computer does not, but it seems to be a matter of degree.

As Blackburn notes, it is much more difficult for a computer to pass a Turing test of any complexity than it was thought it would be in the early days of computer science and AI.

Again from Wikipedia,



I guess I am more sympathetic to the Strong AI position, but I’m very doubtful we’ve seen any technology that would approach being usefully described in those terms.

It isn't merely a matter of simulating the information. You'd have to simulate the chemistry as well, including some slightly more complicated organic systems. But why would we want to run that experiment? Why would we want to create an intelligence that could know true fear, only to have it panic when it finds itself a man-made creation without a body? Perhaps we would want to create one with only warm orgasmic feelings. But I don't think our AI's should be created in such a way. If what we seek is the processing of information, then we should leave out emotions. Emotions are an evolutionary construct meant to guide us through tribal life. The same can be accomplished with operating rules. Informational only, no pain but also no pleasure.

Such an intelligence, if self-aware, would be alien to us at the level of emergent phenomenon. Meaning, we would understand how to build it and would have its blueprints, but we could not close our eyes and imagine what it's conscious feels like, subjectively.

I don't think we really have a choice, we're not about to collectively decide not to pursue more and more advanced technology.

http://en.wikipedia.org/wiki/Friendly_a ... telligence