Hacker News Reader

Can you think of a test or empirical observation that would convince you that a model does "actual reasoning"?

Do bear in mind the tests that very smart people have proposed in the past (from playing chess to holding a conversation to understanding pictures). And consider the implication for your position if you find it hard to devise a well defined test that would convince you to abandon this position.

View on HN

How I'd do it:

1. Define reasoning (hard)

2. Inspect the system to see if this is what it's doing (hard)

Conclusion: too hard, so until I get better I don't know if any given thing is reasoning.

What "very smart people have proposed":

1. Look at something I've only seen people-I-assume-to-be reasoning do

2. Run the system and see if it can do this

3. If satisfied or in the dictionary business, redefine reasoning

4. goto 1

Reply on news.ycombinator.com