Mindforth

jmdrake jmdrake_98 at yahoo.com
Fri Dec 6 11:28:34 EST 2002


Erik Max Francis <max at alcyone.com> wrote in message news:<3DED8D83.EBF1663 at alcyone.com>...
> spurk wrote:
> 
> > No thank YOU for contributing your work to the community.  I've read
> > much of theory before I downloaded forth and mind.f, and I was and am
> > very impressed
> 
> I'm not sure whether it's more depressing or more frightening for people
> to get sucked into this.  Depressing since they should know better or
> frightening because ... they should know better.

Hello Erik,

While I've yet to be able to do anything useful with Mindforth I would
strongly disagree with your assertion that it is "depressing" that 
someone would get "sucked into" it.  First of all Authur isn't charging
anyone anything for this.  Even the e-book is freely available.  So
the most that anyone would be out of for pursuing this would be time.
Now before you go with the "time is precious" argument, is spending
time coding an A.I. program that may never produce any results any
more of a waste than say playing Quake III Arena? :-) (I like computer
games too.)

To Authur,

If you wish to be taken seriously then you need to aim lower on the
theory and aim higher on the application.  What do I mean by this?
Well I think coding you AI and releasing it to the public is a good
first step, but you've really not given much information on how
someone could actually USE it in it's PRESENT stage.  Even your
e-book was of no help in this area.  The "excercises" given basically
amount to "go off and do something with this system that I haven't
been able to accomplish yet" as opposed to "Try this experiment and
see if you get the same results" type of excercise.  Frankly I'm
still stuck on "cats eat fish".  And I seem to recall that this work
better a couple of years ago (when I last looked at it) than it
does now.  At that point it seemed to work as a somewhat quirky
associative database.  Now the JavaScript version seems to go
off on wild tangents spitting out nonsense faster than I can
type in "concepts".  And as has been mentioned the Forth version
seems to have a bug.  Are you able to run the Forth version in
it's current state?  If so which version of Forth are you using?
Again the problem seems to be the AI is in some sort of "race
condition".  If you keep typing fast enough you can delay MindForth
from crashing.  I think the JavaScript version really has the
same bug in it.  At some point I get a message saying that "this
script will run slowly, do you wish to stop it".  I think it
would be better if the AI didn't try to talk while the user is
entering data.  Perhaps a "shut up and listen" mode.

Regards,

John M. Drake



More information about the Neur-sci mailing list