"Searle understands nothing of Chinese, and yet, by following the program for manipulating symbols and numerals just as a computer does, he sends appropriate strings of Chinese characters back out under the door, and this leads those outside to mistakenly suppose there is a Chinese speaker in the room. "
I have trouble with this too. I think it's actually incorrect, or at least misleading. I think what it's _trying_ to say is that even if an entity can perform a complex task doesn't mean it can understand a complex task.
I think the more important result of this argument is that certain complex tasks can be "pre-baked" into rulesets _by an existing intelligence_. To me this just means that intelligent entities can sort of copy parts of their intelligence into other entities which are not intelligent i.e. computer programming.
I think with this argument they're trying to say "a series of sufficiently complex if statements isn't necessarily intelligent" by choosing something we know computers are good at - string manipulations and applying it to something we consider intelligence - language translation.
The argument holds that the computer is obviously not intelligent because it's just a function that takes a character and outputs another character.
But it needs to be a convincing translation, right? The computer would then be able to spit out not just accurate translations but also properly converted cultural idioms and new combinations of words where one didn't exist in the other language. That requires context of surrounding characters, memory of common language use, statistical analysis and creativity.
One implication that arises from this argument is actually about humans. How do we know that we aren't all just incredibly detailed rulesets ourselves without any actual understanding?
Well, first off - we technically can't prove it for anyone other than ourselves. More pragmatically, it's obvious that we, unlike the computer translator, can probe ourselves and be probed by others on whether or not we understand the subject. It's not like we're a bunch of Boltzmann's Brains that just happened into existence. We evolved intelligence in order to survive, not to "trick" other intelligent beings into thinking we're more intelligent than we are. There's no need for that. There's no one smarter around that we need to "trick".
I have trouble with this too. I think it's actually incorrect, or at least misleading. I think what it's _trying_ to say is that even if an entity can perform a complex task doesn't mean it can understand a complex task.
I think the more important result of this argument is that certain complex tasks can be "pre-baked" into rulesets _by an existing intelligence_. To me this just means that intelligent entities can sort of copy parts of their intelligence into other entities which are not intelligent i.e. computer programming.
I think with this argument they're trying to say "a series of sufficiently complex if statements isn't necessarily intelligent" by choosing something we know computers are good at - string manipulations and applying it to something we consider intelligence - language translation.
The argument holds that the computer is obviously not intelligent because it's just a function that takes a character and outputs another character.
But it needs to be a convincing translation, right? The computer would then be able to spit out not just accurate translations but also properly converted cultural idioms and new combinations of words where one didn't exist in the other language. That requires context of surrounding characters, memory of common language use, statistical analysis and creativity.
One implication that arises from this argument is actually about humans. How do we know that we aren't all just incredibly detailed rulesets ourselves without any actual understanding?
Well, first off - we technically can't prove it for anyone other than ourselves. More pragmatically, it's obvious that we, unlike the computer translator, can probe ourselves and be probed by others on whether or not we understand the subject. It's not like we're a bunch of Boltzmann's Brains that just happened into existence. We evolved intelligence in order to survive, not to "trick" other intelligent beings into thinking we're more intelligent than we are. There's no need for that. There's no one smarter around that we need to "trick".