Does Alphago capture human intuition?

djhbrown · Post by **djhbrown** » Tue Oct 10, 2017 6:30 pm

It has been claimed that Alphago captures human intuition [1,2]. The claimed justification for the claim seems to be that (a) artificial neural nets learn like human neural nets learn, and (b) sometimes, pros can't explain why they chose a particular move.

(b) is demonstrably true.
re (a), both artificial and natural neurons make binary decisions [3]; however, the conventional wisdom is that biological neurons are analogue, not digital.

1. https://www.theverge.com/2016/3/10/1119 ... eepmind-ai
2. https://www.theatlantic.com/technology/ ... on/476508/
3. http://sites.google.com/site/djhbrown/H ... ects=0&d=1, pp125-128.

Gomoto · Post by **Gomoto** » Tue Oct 10, 2017 9:45 pm

Why do you ask?

I have a clear intuition in this case

djhbrown · Post by **djhbrown** » Tue Oct 10, 2017 11:55 pm

Gomoto wrote:Why do you ask? I have a clear intuition in this case

Do you say you have a clear intuition in this case for some special reason ?

Gomoto · Post by **Gomoto** » Wed Oct 11, 2017 7:25 am

Don't you FEEL it when you experience the neural network

No special REASON

one of a kind, the network and your mind

Mike Novack · Post by **Mike Novack** » Wed Oct 11, 2017 7:29 am

djhbrown wrote: .... however, the conventional wisdom is that biological neurons are analogue, not digital.

That, however, is more theoretical than practical. As the "sampling rate" (or fineness of division) of the digital goes up, the difference between that and true analog decreases without limit.

We listen to digitally encoded music, watch digitally encoded moving images, etc. without noticing that not true analog << the complaints of vinyl buffs are really about the standard sampling rate used. At some high enough sampling rate the most experienced audiophile would NOT be able to detect a difference >>

Gomoto · Post by **Gomoto** » Wed Oct 11, 2017 7:31 am

quantum world anyone?

Life In 19x19