Artificial Intelligence and Machine Learning

Artificial Intelligence and Machine Learning

Monday: artificial intelligence (AI), turing test, Watson.
Wednesday: machine learning, classification, spam filtering.

Discussion Questions:

Monday
  • Why is Watson a big step for AI?
  • What difficult computational problems had to be solved to make Watson good at Jeopardy?
  • Besides playing game shows, what might the Watson technology be opening the door for?
  • What is the Turing test? (According to Christian? According to Turing?)
  • Why is it good that computers cannot today pass the Turing test?
  • Which of Turing’s predictions from 1950 have come true? Which have not?
  • Why does Turing go into such specifics on what a computer is?
  • Is it true that computers can only do what they are programmed to do? (Compare to Lady Lovelace’s Objection.)
  • What is the importance of universality in Turing’s argument?
Wednesday
  • What is the difference between supervised learning, unsupervised learning, and reinforcement learning.
  • The title of Abu-Mostafa’s article is a bit misleading, it is about machine learning not whether or not computers can think. Compare the topics discussed to those of Turing to the difference between machine learning and computers that can think.
  • Compare Nexflix’s problem of recommending movies to Facebook’s problem of choosing posts for your news feed. What are the similarities or differences?
  • What is an A/B test? How do Facebook’s A/B tests compare to their infamous mood study. What are the key similarities and differences?

Readings and Media:

Monday:

Wednesday

XKCD: AI