Read chapters 1 and 2 of Probably Approximately Correct by Leslie Valiant and answer the following questions.
- What does Valiant mean by theoryful and theoryless? In your view, is linguistics theoryful or theoryless?
- Why does Valiant view computation as essential to life?
- In Valiant’s view, what phenomenon is central to learning? Why?
- Why does Valiant believe that mathematics is key to understanding learning?
- What is an ecorithm and how does it differ from an algorithm? Do you think the term is justified?
- Valiant discusses predicting the 2008 financial crisis. What are some of the reasons he gives for why it may be understandable why the crisis was not predicted? What are some of the reasons you think it would be reasonable for a learning algorithm to fail to learn its target?
- The second chapter mostly lays out Valiant’s ambition to “[treat] Darwinian evolution as a computational learning mechanism and quantitatively [analyze] its behavior” in order to “resolve the central scientific question of whether some instance of natural selection does fit the constraints that have ruled in this universe.” This is a very ambitious program, and it will not be a focus of our class. (No question here)
- Please provide any additional reflections you have on these first two chapters. If you have questions, please include them here.