I don;t understand why people have any respect at all for Searle's"argument", it's just a bare assertion "machines's can't think", combined with some cheap misdirection. Can anyoneone argue that having Chinese characters instead of bits going in and out is something other than misdirection? Can anyone argue that having a human being acting like a cpu instead of having an actual cpu is something other than cheap misdorection?
I think you might be missing out on what the Chinese Room thought experiment is about.
The argument isn’t about whether machines can think, but about whether computation alone can generate understanding.
It shows that syntax (in this case, the formal manipulation of symbols) is insufficient for semantics, or genuine meaning. That means whether you're a machine or human being, I can teach you every grammatical rule or syntactical rule of a language but that is not enough for you to understand what is being said or have meaning arise, just like in his thought experiment. From the outside it looks like you understand, but the agent in the room has no clue what meaning is being imparted. You cannot derive semantics from syntax.
Searle is highlighting a limitation for computationalism and the idea of 'Strong AI'. No matter how sophisticated you make your machine it will never be able to achieve genuine understanding, intentionality, or consciousness because it operates purely through syntactic processes.
This has implications beyond the thought experiment, for example, this idea has impacted Philosophy of Language, Linguistics, AI and ML, Epistemology, and Cognitive Science. To boil it down, one major implication is that we lack a rock-solid understanding or theory of how semantics arises, whether in machines or humans.
I don;t understand why people have any respect at all for Searle's"argument", it's just a bare assertion "machines's can't think", combined with some cheap misdirection. Can anyoneone argue that having Chinese characters instead of bits going in and out is something other than misdirection? Can anyone argue that having a human being acting like a cpu instead of having an actual cpu is something other than cheap misdorection?
I think you might be missing out on what the Chinese Room thought experiment is about.
The argument isn’t about whether machines can think, but about whether computation alone can generate understanding.
It shows that syntax (in this case, the formal manipulation of symbols) is insufficient for semantics, or genuine meaning. That means whether you're a machine or human being, I can teach you every grammatical rule or syntactical rule of a language but that is not enough for you to understand what is being said or have meaning arise, just like in his thought experiment. From the outside it looks like you understand, but the agent in the room has no clue what meaning is being imparted. You cannot derive semantics from syntax.
Searle is highlighting a limitation for computationalism and the idea of 'Strong AI'. No matter how sophisticated you make your machine it will never be able to achieve genuine understanding, intentionality, or consciousness because it operates purely through syntactic processes.
This has implications beyond the thought experiment, for example, this idea has impacted Philosophy of Language, Linguistics, AI and ML, Epistemology, and Cognitive Science. To boil it down, one major implication is that we lack a rock-solid understanding or theory of how semantics arises, whether in machines or humans.
I agree. At it's heart, it just relies on mysticism. There's a hidden assertion that humans are supernatural.
First time I heard of this was in Blindsight, and every time I use an LLM it just makes me think of the crew talking to Rorschach.
Intelligence without consciousness...
https://www.rifters.com/real/Blindsight.htm