Critique of computationalism
No description yet. This curated deck is ready for daily study.
Card preview (24)
T2-3-4 indistinguishability
The Turing indistinguishability: T4: Symbolic, sensorimotor and neuromolecular indistinguishability T3: Symbolic and sensorimotor indistinguishability T2: Symbolic indistinguishability
Strong AI
Behaves like humans
Weak AI
Has some human like characteristics
Symbol grounding problem
Just manipulating the symbols is not by itself enough to guarantee cognition, perception, understanding, thinking and so forth. Insisting on implementation independence, means that you can only design T2-indistinguishable systems, which will not have intrinsic semantics and thus not reflect cognition, which seems to be about something. To get semantics, you need at least T3-indistinguishability, losing implementation independence
Chinese room
He is a human being in a room and communicating but he does not know what the Chinese symbols mean
Semantics ≠ syntax
Syntax does not give semantics
Serial processing
Everything happens after each other
Hardware
The physical thing
Software
The inner thing
Searle’s 3 axioms
Axiom 1: Computer programs are formal (syntactic) Axiom 2: Human minds have mental contents (semantics) Axiom 3: Syntax by itself is neither constitutive of nor sufficient for semantics
Searle’s conclusion
Programs are neither constitutive of nor sufficient for minds
Robot argument/reply
If the computer program was causally hooked up with the rest of the world, then it would acquire semantics (it would be able to ground symbols by causally interacting with the world) Searles reply: Imagining the Chinese room inside the robot does not add semantics, even if the robot interacts with the world
Semantics
The meaning of a word, phrase, or text.
Syntax
The arrangement of words and phrases to create well-formed sentences in a language. The structure of statements in a computer language.
Intentionalism
Mental states are about something
Ground (verb)
To provide basis for
The system reply
The room understands it Compare with a single (language) neuron relative to the whole brain Searles reply: Saying that the room understands Chinese is analogous to claiming that he (Searle) understands Chinese, if he memorizes all the symbols and all the rules
The brain simulator reply
If the operations of the brain of a Chinese speaker were simulated, such a system would understand Chinese, just like any Chinese speaker does We build something that is very similar to the brain structure- If we can do this there must be understanding Searles reply: (Lau not sure what Searle means) “Computer simulations of brain processes provide models of the formal aspects of these processes. But […] [t]he computational model is no more real than the computational model of any other natural phenomenon”
Church/Turing Thesis (CTT)
All independent attempts to formalize what mathematicians mean by a “computation” or an “effective procedure,” even when they have looked different on the surface, have turned out to be equivalent
Semantic operations
Operations that are based on the meanings of symbols, rather than just their shapes
Trivial computation
Generating uninterpretable symbol systems
Nontrivial computation
The kind that can be made systematic sense of. These are the ones we are interested in
Weak CTTP
Claims that every physical system is formally equivalent to a Turing Machine
Strong CTTP
A plane is a computer It is wrong or uninteresting If every physical process is in fact computation, then the cognitive scientist might as well have ignored the computationist’s message and just kept on hunting for what the right physical process(es) might turn out to be.