Agree much w your take on DD. But his trick is to introduce syntax and semantics = transduction. His illusion and trick for his audience. These interfaces are alike but not =. Unclear that he successfully responded to Searle.
Hah, I agree that his response to Searle in "The Intentional Stance" fell short. FWIW, Brian Cantwell Smith did a much better job on that in "On the Origin of Objects." As far as syntax + semantics vs. transduction, I rather like Maturana and Varela's view of the nervous system acting to maintain internal relations between the organism's sensors and effectors.
"Imagine a person who has always lived in a submarine. He has never left it and has been trained how to handle it. Now, we are standing on the shore and see the submarine gracefully surfacing. We then get on the radio and tell the navigator inside: “Congratulations! You have avoided the reefs and surfaced beautifully. You really know how to handle a submarine.” The navigator in the submarine, however, is perplexed: “What’s this about reefs and surfacing? All I did was push some levers and turn knobs and make certain relationships between indicators as I operated the levers and knobs. It was all done in a prescribed sequence which I’m used to. I didn’t do any special maneuver, and on top of that, you talk to me about a submarine. You must be kidding!”
All that exists for the man inside the submarine are indicator readings, their transitions, and ways of obtaining specific relations between them. It is only for us on the outside, who see how relations change between the submarine and its environment, that the submarine’s behavior exists and that it appears more or less adequate according to the consequences. If we are to maintain logical accounting, we must not confuse the operations of the submarine itself and its dynamics of different states with its movements and changing position in the environment." (H. Maturana and F. Varela, "The Tree of Knowledge: The Biological Roots of Human Understanding")
Karl Friston is also working along these (dialectical ?) lines in spectacular ways. Andy Clark playing Huxley to Friston’s Darwin. Clark also a Dennett enthusiast re: illusions of consciousness. Geoff Hinton aligned here too via neural nets.
Nice tribute to a very broad thinker. Re: Habermas, is the concern that A.I./LLMs are not accountable for their speech acts in the same way humans are? Or something else?
It has more to do with the fact that humans use speech acts to communicate ethical and moral precepts, and moreover they are aware that this is what is taking place. We can distinguish our perception of speech acts from other forms of perception. AI systems do not possess the ability to make such distinctions, but humans will be tempted to project them onto the linguistic outputs generated by AI systems.
Agree much w your take on DD. But his trick is to introduce syntax and semantics = transduction. His illusion and trick for his audience. These interfaces are alike but not =. Unclear that he successfully responded to Searle.
Hah, I agree that his response to Searle in "The Intentional Stance" fell short. FWIW, Brian Cantwell Smith did a much better job on that in "On the Origin of Objects." As far as syntax + semantics vs. transduction, I rather like Maturana and Varela's view of the nervous system acting to maintain internal relations between the organism's sensors and effectors.
"Imagine a person who has always lived in a submarine. He has never left it and has been trained how to handle it. Now, we are standing on the shore and see the submarine gracefully surfacing. We then get on the radio and tell the navigator inside: “Congratulations! You have avoided the reefs and surfaced beautifully. You really know how to handle a submarine.” The navigator in the submarine, however, is perplexed: “What’s this about reefs and surfacing? All I did was push some levers and turn knobs and make certain relationships between indicators as I operated the levers and knobs. It was all done in a prescribed sequence which I’m used to. I didn’t do any special maneuver, and on top of that, you talk to me about a submarine. You must be kidding!”
All that exists for the man inside the submarine are indicator readings, their transitions, and ways of obtaining specific relations between them. It is only for us on the outside, who see how relations change between the submarine and its environment, that the submarine’s behavior exists and that it appears more or less adequate according to the consequences. If we are to maintain logical accounting, we must not confuse the operations of the submarine itself and its dynamics of different states with its movements and changing position in the environment." (H. Maturana and F. Varela, "The Tree of Knowledge: The Biological Roots of Human Understanding")
Karl Friston is also working along these (dialectical ?) lines in spectacular ways. Andy Clark playing Huxley to Friston’s Darwin. Clark also a Dennett enthusiast re: illusions of consciousness. Geoff Hinton aligned here too via neural nets.
Nice tribute to a very broad thinker. Re: Habermas, is the concern that A.I./LLMs are not accountable for their speech acts in the same way humans are? Or something else?
It has more to do with the fact that humans use speech acts to communicate ethical and moral precepts, and moreover they are aware that this is what is taking place. We can distinguish our perception of speech acts from other forms of perception. AI systems do not possess the ability to make such distinctions, but humans will be tempted to project them onto the linguistic outputs generated by AI systems.