Hi,

Simply put, I would like to see the wikipedia listing for Hierarchical
Temporal Memory improved and brought up to date.

Problems:
The current listing has several inaccuracies regarding the current state of
HTM Theory; does not reflect the historical evolution of the theory; and
instead, superimposes the historical/older version's origins with newer
developments creating a cognitive mishmash of indistinct informational
attributes.

Solution:
I would like to see the current page/listing accurately describe the events
following the publishing of "On Intelligence"; the formation of Numenta;
and the eventual split between approaches favored by Dileep George, and
Jeff Hawkins. I propose that coverage of Dileep George and his company
Vicarious link to another page that those in that camp can more fully
enhance at their will - and the rest of the listing then proceed to
describe the development of the Cortical Learning Algorithm and the more
current focus of Numenta et al.

Unknowns:
1. Is the term "Hierarchical Temporal Memory" currently used by Dileep
George and Vicarious?

Process:
Once this proposal has been honed such that it reflects a satisfactory
approach and direction; that the current listing text be incrementally and
iteratively modified, then submitted for approval - each submission's
approval representing an interactive improvement of the entire text until
such time as it is considered to be an accurate representation. This entire
process is to proceed "offline" (not posted to wikipedia until finished).

I propose this iterative approach because it will make things easier
(initially for me specifically, but maybe by a "team" of people who would
like to participate - I encourage this). Advantages:

1. the text updates can be made in small increments as knowledge is
acquired seeing as the "implementing team" doesn't possess the whole story.
[ Again, the "implementing team" will be initially composed of myself but
is wide open to other community contributors ;-) ]

2. Submissions of small changes won't take much time to "validate" and turn
around time can be more "immediate" leading to a quicker process.This will
consume negligible time on the part of the  validator, increasing
willingness and therefore the process' general momentum.

What does everyone think?

Regards,
David

-- 
*With kind regards,*

David Ray
Java Solutions Architect

*Cortical.io <http://cortical.io/>*
Sponsor of:  HTM.java <https://github.com/numenta/htm.java>

[email protected]
http://cortical.io

Reply via email to