A good general resource to start reading on the topic:
Stanford Science in the News

Turning, one of the greatest of our minds, on the idea of intelligence machines:
Computing Machinery and Intelligence (1950)
Turning proposes the classic imitation game (also called the Turning Test). Further, he provides nine (9) objections to the use of said test.

1. The Theological Objection:
       Concept: Thinking is a function os a man’s soul, and machines do not have a soul.     Refute: Such a concept is rooted in noma. But, playing to the Theologian realm, what is to stop God from allowing man to build the vessel and to fill it with a soul himself?
2. The “head’s in the sands” Objection:
       Concept: It is too scary to consider, let’s hope it is not possible.     Refute: The prospect is just as exciting as it is scary. There is no reason here to disregard the test.
3. Mathematical Objection:
       Concept: Referring back to Godel’s theorems and Church/Turning’s theories of machines, if the machine is built on a set of rules, then it is necessary that such machine cannot correctly answer all questions.
       Refute: Why must the answer be correct? Who is to say that the same limitations do not applying to human intelligence?
4. The Argument from Consciousness:
       Concept: Intelligence is based in consciousness and consciousness is defined by the thoughts and emotions.
       Refute: Since we can only tell if someone is feeling something as we do, we cannot assert intelligence on anyone else. As such, it is not a viable point of analysis.
5. Argument from Various Disabilities:
       Concept: There are certain things that a machine cannot do, these define where they cannot be intelligent.
       Refute: This critique falls into a lack of creativity (much of which was rom the computational limits at the time).
6. Lady Lovelace’s Objection:
       Concept: Machines can do only what we tell them to do.
       Refute: We might be telling them the minute actions to take, but the resulting emergent behavior is beyond what we “tell them to do.”
7. Argument from Continuity in the Nervous System:
       Concept: The real word is not built off discrete-state machines.
       Refute: If the machine is adequately built it will self correct for any short comings when it comes to being restricted to a discrete space.
8. The Argument from Informality of Behavior:
       Concept: Man can fill in the gaps when rules are not defined. Since machines are defined off rigid logic systems, then they won’t be able to account in the same way.
       Refute: The initial premise stated about man is unfounded and could be false.
9. The Argument from Extrasensory Perception:
       Concept: What is ESP is true and human are able to use it to effectively tip the scales?
       Refute: There is no considerable defense for ESP, but even if there were the bias could go both directions (clairvoyance on the part of the human and psychokinesis on the part of the investigator). Either way, the variable could likely be controlled for in the test. After all of this, Turing offer how such an intelligent system might be built. He first outlines the idea of creating a massive knowledge base, encoding the whole of the Encyclopedia. This would be like building a “adult machine”, instead we could build a “child machine” and teach it what it must learn. With this, he introduces the idea of an evolutionary process that is directed (similar to a kind of differential evolution; but that is heavily adding labels to the idea). The idea of “punishments and rewards” is brought up as a method to teach the system. This should be recognizable as the fundamental ideas that outline reinforcement learning and supervised learning. Finally, there is an important idea about the importance of emergent knowledge:
   An important feature of a learning machine is that its teacher will often be very largely ignorant of quite what is going on inside, although he may still be able to some extent to predict his pupil's behavior.