[ExI] Neural networks score higher than humans in reading and comprehension test

[spike]

From: extropy-chat [mailto:extropy-chat-bounces at lists.extropy.org] On Behalf Of Dylan Distasio
Subject: Re: [ExI] Neural networks score higher than humans in reading and comprehension test

 

>…This is the Achilles heel of all the deep learning that I've been exposed to.  Deep nets are also great at image recognition until an adversarial attack is injected that will fool a net, but not a human child…

 

Ja, this is all Stanford’s fault.  I bet I could write a test that would fool that evil computer and those wily programmers.

 

 

>…There is zero comprehension.  The nets are dead behind the eyes…

 

Sooner or later, we are going to figure out how our own comprehension works, and model it mathematically.  When that happens, we too are dead behind the eyes.  

 

Hey, it happened in chess and go.

 

 

>…In the mean time, this will become more and more of a problem as these systems get embedded in our environment and become open to exploit for profit:

 

>…https://blog.openai.com/adversarial-example-research/

 

Wait what?  Exploiting for profit is a problem?  I thought that was the solution.

 

spike

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20180116/f620bcaf/attachment.html>