Question

Criticize Turing’s criteria for computer software being intelligent.

Answer 1

Answer:

$\huge\pink{\mid{\underline {\overline{\tt answer}}\mid}}$

1. It is highly anthropocentric. Essentially: “If you don’t behave like a human, you can’t be intelligent.” (Or, at least, you have to be able to mimic a human. Can we not conceive of a type of intelligence which is alien to us, but not so advanced it can mimic humans?)
2. The Chinese Room argument. If something walks like a duck, and talks like a duck, it might be a plastic duck toy, it’s not necessarily a duck. You can get exactly the same behavior without any underlying intelligence by programming the right responses.
3. The value of the test depends on the judge conducting the experiment. Some people are tricked more easily than others. The value of the test also depends on how well humans actually can be expected to do. Humans are far from rational beings, with some built-in and more culturally developed biases. This reminds me of an episode of “Agents of SHIELD” where they are trying to determine whether people are real people or “Chronicoms”, a sort of time-traveling android. One of the tests is checking their reaction to the word “moist”. If they do not cringe at least a little, they might not be human — this is based on the bias that many people have against that particular word. I liked the episode and the choice of tests. Instead of the usual questions people might think of (test their knowledge of history, see how rational they are, …) it tested for what can be considered flaws in humans, instead.

$\huge\pink{hope\:it\:helps\: you\:}$

$\huge\pink{Mark\:me\:as\: brainlist\:}$