Reading Guide: George Luger, "AI: Early History and Applications"

Some of these questions may not be so answerable until after you've read the whole reading...

1.1

What is the connection between Mary Shelley and Frankenein and artificial intelligence?

Explain how Aristotle's distinction between matter and form was crucial to the early development of AI. (Do any AI researchers reject this distinction?)

How has our notion of 'mind' developed throughout history? How does this relate to AI?

Do you agree with Pascal, that his "arithmetical machine produces effects which approach nearer to thought than all the actions of animals."?

Are mind and body distinct, like Descartes said? Prove your answer.

Hume said (very roughly) that knowledge is just experience built from repeated associations. If we believe that, is it more or less likely that we'll be able to construct intelligent machines? (Note that I'm being slippery with my terminology.)

What does the author mean by "graph"?

Why did early logicians (including Aristotle) want to study logic?

Explain the significance of logic to AI.

What is the relationship between intelligence and AI? (Or part of the relationship, anyway.)

So we have machines that could, potentially, be intelligent. What's tricky about deciding whether they actually are? That is, what problem is the Turing Test trying to solve?

But why doesn't everyone love the Turing Test?

Is there a connection between culture and intelligence? What does your answer imply for AI?

What did Wittgenstein and Heidegger say that was so different from what had gone before? (Be brief.)

How did AI research respond to these ideas?

1.2

What's nice about using games for AI research? What are the limitations of this approach?

What is "state space search," "search space," etc., talking about?

What's a heuristic?

What's a theorem? What is "automated theorem proving"?

What is an "expert system"? Does it make sense to call a computer an expert?

What is "domain-specific knowledge"? What are the advantages and disadvantages of using only this type of knowledge?

Does "natural language understanding" require anything beyond understanding rules of language? Explain.

What might it mean for a machine to "learn"?

[Reflect on the author's 12 questions at the end of the reading.]