Yesterday's post was quite long and wide-ranging, with a lot of examples etc. It focused specifically on how it might be possible to get semantics from syntax, particularly with reference to The Chinese Room thought experiment.
However I feel that I could sum up my position more succinctly by looking at the meaning of the word "meaning" itself.
The question is where does meaning come from, and whether it is possible for meaning to be inherent within a purely formal system such a computer program without the need for some external mind to infer it.
But what does "meaning" mean?
I would argue that for something to have meaning, there has to be a mind that understands that meaning.
If something in the external world is to have meaning, the process of understanding that meaning would be taking this perceived object or percept (e.g. a word, a symbol or a statement), and mapping it to concepts within the mind. Something has meaning if it is possible for a mind to do this -- if there is a well defined way to translate this external percept to a mental representation.
Thoughts within the mind have inherent meaning because they already exist as mental representations. In this view, the term "meaningless thought" is an oxymoron.
But if meaning is simply the ability to form mental representations, then there is no problem in supposing that artificial intelligences could have meaning. Artificial intelligences would have ways of representing concepts in their minds just as we do. Their thoughts would inherently have meaning to them even if this meaning is not obvious to an external observer. This is entirely analogous to the difficulty in identifying the meaning of thoughts in a human brain by analysing a functional MRI scan.
For John Searle and other deniers of Strong AI, meaning is mystical and ineffable. They see no way such a magical property could be attributed to a machine.
For me, the concept of "meaning", like so many concepts in philosophy, is either meaningless (!), ill-defined or entirely trivial. There is no mystery.
However I feel that I could sum up my position more succinctly by looking at the meaning of the word "meaning" itself.
The question is where does meaning come from, and whether it is possible for meaning to be inherent within a purely formal system such a computer program without the need for some external mind to infer it.
But what does "meaning" mean?
I would argue that for something to have meaning, there has to be a mind that understands that meaning.
If something in the external world is to have meaning, the process of understanding that meaning would be taking this perceived object or percept (e.g. a word, a symbol or a statement), and mapping it to concepts within the mind. Something has meaning if it is possible for a mind to do this -- if there is a well defined way to translate this external percept to a mental representation.
Thoughts within the mind have inherent meaning because they already exist as mental representations. In this view, the term "meaningless thought" is an oxymoron.
But if meaning is simply the ability to form mental representations, then there is no problem in supposing that artificial intelligences could have meaning. Artificial intelligences would have ways of representing concepts in their minds just as we do. Their thoughts would inherently have meaning to them even if this meaning is not obvious to an external observer. This is entirely analogous to the difficulty in identifying the meaning of thoughts in a human brain by analysing a functional MRI scan.
For John Searle and other deniers of Strong AI, meaning is mystical and ineffable. They see no way such a magical property could be attributed to a machine.
For me, the concept of "meaning", like so many concepts in philosophy, is either meaningless (!), ill-defined or entirely trivial. There is no mystery.
No comments:
Post a Comment