Mike Archibold said:
It seems like the mega brain research projects get 10 years (BRAIN
initiative and the HBP) -- somebody else is paying the bills.

But the independent researcher/developer is paying his own way -- so
it's a lot faster schedule, one or two years!  ;)
------------------
 
I guess that there is some truth to that.  However, my feeling is that I have a 
simple model that I can use to get a jump start on the program.  The mega-brain 
research projects are starting at the bottom (as in bottom up).  The simplicity 
of my starting model means that (if my theories are good then) as I encounter 
problems I should be able to improve the algorithms with feasible simple 
modifications.  Currently, most AI/AGI algorithms work really well in some 
cases but not very well in other cases.  I expect that if my theories are 
reasonable I will run into the same problem.  But, because I have my simple 
model for AGI is central to AGI then I think I may be able to improve on the 
model gradually.  I would expect the diminishing returns problem, but I can 
live with that.  My criticism of contemporary central paradigms are that they 
are paradigms of narrow AI and all the AGI stuff is all about going crazy over 
getting narrow AI algorithms to act like they are AGI.  Of course it can be 
done, but if you can find some simple insights into the nature of the thing and 
those insights are technologically feasible, then that is a much better way to 
get started. 
 
OK, so there might be a bit of manure in there.  But if you know anything about 
growing plants then you know that fertilizer is an important part of the 
process.
 
I thought I came up with a compelling argument about why the use of a discrete 
model would be better than using a weighted model to advance the field right 
now.  I realized that no one would be convinced but it might make some people 
think.  Then as I thought about it, I realized that it was not only a 
compelling argument, but that it might explain a lot.
 
I am not going to tell you what that argument was (because I already have 
talked about the parts of the argument) but I will give you a hint.  I have 
been stuck for years thinking about the recognition problem.  But look what we 
do in this group.  We are not sweating over our problems recognizing something 
but on getting other people to recognize what we are talking about.   (OK, so 
we feel comfortable about interpreting what is going on so that may not seem 
like a big problem for us.)  But, none the less, I have been looking at the AGI 
problem as a pure recognition problem without seeing it as an expression 
problem.  I used expression in my theories (like imaginatively projecting 
previously acquired knowledge onto a situation in order to compare a conjecture 
about the situation with the situation) but notice that even this leaves the 
fundamental expression problem out.  Most of my effort is: How do I get other 
people to understand what I am talking about?
 
Jim Bromer 
 
 
 
Mike Archibold said:
It seems like the mega brain research projects get 10 years (BRAIN
initiative and the HBP) -- somebody else is paying the bills.

But the independent researcher/developer is paying his own way -- so
it's a lot faster schedule, one or two years!  ;)

On 5/27/13, Jim Bromer <[email protected]> wrote:
>
> So I will say that
> if my idea is right I should definitely have something to show within 2
> years.
                                          


-------------------------------------------
AGI
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-f452e424
Modify Your Subscription: 
https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-58d57657
Powered by Listbox: http://www.listbox.com

Reply via email to