Live data from Hacker News

Ask HN: What would you ask Peter Norvig?

news.ycombinator.com

1–5 of 5 posts

Re: Ask HN: What would you ask Peter Norvig?

#2
Who has been the single most influential person in your life?

Kind of a cliche question, and perhaps not one worth asking, but I've heard it asked of some pretty impressive people and the answer has always been really fascinating. Probably not worth it though, if you only have one question, especially if you're looking for something more technical.

Re: Ask HN: What would you ask Peter Norvig?

#5
I really admire Peter Norvig and have learned a lot simply by reading (and re-reading) two of his essays: How to Write a Spelling Corrector[1] and Solving Every Sudoku Puzzle[2].

What fascinates me about these essays is that in one, he is able to get good results from using a simple algorithm over a large corpus of data (books from Project Gutenberg). In the other, he is able to get good results through the application of a clever algorithm (constraint propagation). Both of these approaches (lots of data, clever algorithms) are used in AI. My question would relate to their relative merit, and would be something like the following:

"Have advances in Artificial Intelligence been driven more by the discovery and application of novel algorithms and techniques, or by an increased ability to collect, store, curate and analyze increasing bodies of data?"

However, I would also really like to know how he feels about John Searle's Chinese Room[3] argument.

[1] http://norvig.com/spell-correct.html

[2] http://norvig.com/sudoku.html

[3] http://en.wikipedia.org/wiki/Chinese_room