The Turing Test
test was used to measure artificial intelligence
The ability to carry out a task which would require intelligence if carried out by a human
Explain how the Turing Test was conducted.
A human subject sits at a terminal in one room and communicates with two terminals in another room. One of the other terminals is used by a human. If the human test subject cannot identify which responses are from the computer, then the computer system is said to be “intelligent”.
What are some of the flaws of the Turing Test?
(1) The test can show the appearance of intelligence, but cannot prove its existence (2) No agreed definition of intelligence exists (3) The intelligence of the human is a factor. (4) The test can show the recognition of language, but that is not necessarily proof of intelligence.
You’re placed in front of a wall and given a computer with a monitor and keypad. You are to use the computer and send messages to someone on the other side. You can ask or talk about anything. At the end, you must decide-based on the conversation you just had-whether you have been conversing with a person or computer.
A test proposed by British mathematician Alan Turing, and often taken as a test of whether a computer has human like intelligence.
Purpose of Turing test
Turing thought that if we could build a computer that could pass as a real person under these conditions, then we are in the position to say that this computer can think
What makes Turing a functionalist? Be specific.
he doesn’t care how consciousness seems possible, just that it does. Take the Chinese room for example. Searle doesn’t understand Chinese, but using the manuals, he passes the Turing test for understanding because he appeared to consciously understand the language.
What is the Chinese Room? What is it intended to demonstrate?
an example argument where searle sits in a room and answers questions in Chinese with the help of several manuals. According to the person outside the room asking the questions, searle passes the turing test for understanding of Chinese. This demonstrates flaws in the turing test, because as we can see, searle doesn’t truly understand Chinese.
FIRST objection to turing test
1. the “heads in the sand” objection
– if computers can think, they will be smarter than humans, and take over the earth
BUT, this is appeal to fear
SECOND objection to turing test
2. the objection from consciousness
– thinking is not just a matter of processing information, it involves consciousness and awareness
BUT, could be that certain machines are conscious when processing info
ALSO, if we raise the standars for thinking too high, we encounter the problem of other minds
3. the novelty objection
– computers cannot do anything new: they can only do what they are told to do
Searle’s argument against strong AI:
1. computer programs are formal systems (they are syntactic)
2. human minds have contents, or meanings
3. syntax, by itself, is neither constitutive nor sufficient for meaning