Connectionists: how the brain works?

[james bower]

On Mar 20, 2014, at 2:37 PM, Brian J Mingus <brian.mingus at Colorado.EDU> wrote:

> Having a bunch of people doing whatever they want in parallel is an extremely inefficient way to solve a problem.

on that we agree

> The way to solve large problems most effectively is to have a few architects at the top design the research paradigm as a system of interacting modules.

On that I also agree  - the problem is when using the wrong tools and paradigm.
 

> The architects maintain conceptual integrity and the modules (research programs) can be implemented in parallel. This is not scientific ADD.

big data is - IMHO - 


> This is how we put a man on the moon.

Really, we collected all possible trajectories between Florida and the Sea of Tranquility and then once we had collected them into a huge data base, we set about looking for the best trajectory we could find - under the condition that it didn’t require too many parameters to describe?
Really? I don’t think so.


> Nobody is steering this ship, and the end result is either going to take forever, be a big mess that nobody knows how to make sense of, or an AI that takes over the world. It would be much better to defer the direction of our efforts to a team of our best "ADD thinkers" than it is for us all to pick a random topic, pretend it's the most important thing in the universe and then justify it endlessly in order to make ourselves feel better.

We agree completely, in fact I have written about this for years.  most recently:

https://www.dropbox.com/s/erw705h4yyh3l9k/272602_1_En_5%20copy.pdf

(have posted this before)

Jim


> 
> And with that, I'm out to continue on my merry rational ADD way:)
> 
> $.02
> 
> Brian
> 
> On Thu, Mar 20, 2014 at 1:26 PM, james bower <bower at uthscsa.edu> wrote:
> 	3) the ability to collect massive amounts of data and bang away at it, instead of thinking. (a kind of scientific ADD)

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/connectionists/attachments/20140320/d49ed40b/attachment.html>