Measuring abstract reasoning in neural networks

Re:Measuring abstract reasoning in neural networks by hprotagonist on Sunday July 15, 2018 @11:48AM (#17535710)

This work does not engage with Searle’s Chinese Room argument, as far as I can tell, nor do the authors use any shockingly novel network architecture or advances in approach beyond normal DL approaches.

My position, then, is that the system they describe cannot be said to be reasoning at all.

[ Parent | on HN ]

Re:Measuring abstract reasoning in neural networks by SCHiM on Sunday July 15, 2018 @12:30PM (#17535896)

I think I've never completely understood the Chinese room argument, to me it seems like that whole chain of examples and counter examples are the product of people playing word games and discussing definitions.

If your 'room' 'understands', but no part of the 'room' can be attributed with having 'intelligence', doesn't it stand to reason that the _room_ itself is the intelligent actor? I really don't see the problem with the fact that the human 'cpu' doesn't speak Chinese, the room itself 'understands' Chinese just fine.

[ Parent | on HN ]

Re:Measuring abstract reasoning in neural networks by mindcrime on Monday July 16, 2018 @10:05AM (#17541308)

A lot of people consider Searles' argument to have been refuted quite some time ago. At the very least, I don't see that it deserves any particular status w/r/t determining what is or isn't reasoning.

[ Parent | on HN ]

Re:Measuring abstract reasoning in neural networks by bhouston on Sunday July 15, 2018 @11:59AM (#17535757)

Define reasoning?

Technically in some definitions of reasoning, such as symbolic logic solvers that already exist, those are logical reasoning optimization tools.

[ Parent | on HN ]