Week 1 Flashcards
Why does Turing change the question from whether a machine can “think” to whether a machine is capable of passing the imitation game?
“Think” is an ambiguous term.
Why does Turing insist that the man and the computer be in secluded rooms away from the person who is trying to figure out which is the computer?
It prevents the interrogator from making judgments based on what the computer looks like.
What are examples of Strong AI?
- Makes independent decisions
- Functions like the human brain
- Adapts to new situations on it’s own
What is the conclusion drawn from the Symbol Manipulating Room?
There is only syntactic features without semantic features
Explain the setup of the Imitation Game (or Turing Test) and say what the goal of the test is.
The setup of the Imitation Game or Turing Test is that of an isolation chamber in which all individuals: the investigator, Person A, and Person B, cannot see each other. The investigator needs to figure out which is Person A and which is Person B correctly. In another experimental setup, the investigator needs to find out which person is actually the digital computer in disguise without seeing or hearing both. The investigator will ask each person a question and the responses he will receive will be all typed statements. The goal of this test is to answer the question, “Can digital computers mimic human behavior and/or responses by doing well in the imitation game?” instead of the question, “Can machines think?” The computer needs to convince the investigator that it is a human.
Why does Turing change the question from “Can machines think?” to “Are there imaginable digital computers that can do well in the imitation game?”
Turing changes this question from “Can machines think?” to “Are there imaginable digital computers that can do well in the imitation game?” because the terms “think” and “machines” were both ambiguous. He believed that a more solid definition of both terms would provide a framework in which scientists could actually observe and measure. “Think” could mean anything, Turing believed it had to be more specific such as “human responses or behavior.” “Machines” could also be anything so he focused on “digital computers”.
Explain why the Turing test is a sufficiency test for intelligence.
The Turing test is a sufficiency test for intelligence because it isn’t a necessary condition for an animal/machine to pass. If it does pass the Turing test, then it is intelligent. It just doesn’t mean that the Turing test is the only way to determine whether something has intelligence.
What does it mean to say that the mind is computational (that the mind is like a computer)?
The mind is computational in the sense that we may be algorithm-based creatures. The mind processes information in a mathematical sense that happens too fast for any current technology to measure or calculate, for the time being. With this assumption and/or theory, this means that the mind is separate from the brain. We think of this as the separation of software from hardware, respectively. The mind may be an abstract zap of neural signals that has different kinds of computations, calculations, algorithms, happening all at once. Whereas the brain is just the mechanism that brings the neural signals from all over the body to the mind.
Explain the Symbol Manipulating Room thought experiment and provide the conclusion.
The Symbol Manipulating Room thought experiment involves the setup of a person isolated completely in a room with small openings enough for input and output transactions. The outside person will give information or symbols to the person inside and this person will interpret these symbols based on a preset collection of rules on how to manipulate and translate the input into the desired output. The output is then sent out to the person. They will never know whether or not the person inside understood the meaning of the input symbols but will assume the person inside does based on the accuracy of the output. In conclusion, the syntactic features can be processed without semantics. The SM room has syntactic features but no understanding of the semantics, whatsoever.
What did you think about Turing’s argument that we should accept the Turing test as a means for determining whether a digital computer is intelligent (thinking)? Did you initially think this was a good test? Do you still think that it’s a good way to make the determination?
I think that the Turing test is insufficient to definitely determine whether something is intelligent or not. This is because we are ultimately making the assumption that human behavior/responses is the only way to determine intelligence. There should be other ways of measuring intelligence through other cognitive skills. What about animals such as dolphins and elephants? They cannot communicate like humans or mimic human behavior but have their own ways of solving problems and creating bonds with members of their own communities. I think the Turing test is definitely contributive but it is also severely limited. For the current time, I think it is an okay thought experiment until we can come up with another test for intelligence.
