nifty: startup models human neocortex for memory management

[Giles Bowkett]

http://www.numenta.com/index.php

Founder Jeff Hawkins also founded Palm, Handspring, and a neuroscience
institute which generated the research for this company. There's a
podcast of a speech he made on entrepreneurialism in general at
Stanford here:

http://edcorner.stanford.edu/podcasting.shtml

And somewhere a Real Audio thingy of an interview about his new
company which he did on/for NPR. (I have this idiosyncratic aversion
to RA so I haven't heard it, but the podcast is interesting.)


Giles

[Jim Rutt]

I went out and spent an afternoon with Hawkins at his Redwood Nuerological
Institute not long after the publication of _On Intelligence_ which I had
found intriguing but frustrating in its lack of specificity.   At the time
Hawkins was quite enamored of  (from Numenta web site)

"Dileep George implemented a mathematical formalism of Hawkins' theory,
demonstrating it is possible to express this new type of memory system in
software"

Indeed the MATLAB code was posted on the RNI website.  Alas, I got
distracted on another project and by the time I got back to look at the
code, it had been taken down and Hawkins had spun up Numenta.

I wonder if the code is still available on one of the web archives?

=jim rutt


At 03:08 PM 3/22/2006, you wrote:

>http://www.numenta.com/index.php
>
>Founder Jeff Hawkins also founded Palm, Handspring, and a neuroscience
>institute which generated the research for this company. There's a
>podcast of a speech he made on entrepreneurialism in general at
>Stanford here:
>
>http://edcorner.stanford.edu/podcasting.shtml
>
>And somewhere a Real Audio thingy of an interview about his new
>company which he did on/for NPR. (I have this idiosyncratic aversion
>to RA so I haven't heard it, but the podcast is interesting.)
>
>
>Giles
>
>============================================================
>FRIAM Applied Complexity Group listserv
>Meets Fridays 9a-11:30 at cafe at St. John's College
>lectures, archives, unsubscribe, maps at http://www.friam.org

===================================
Jim Rutt
voice:  505-989-1115

[Jochen Fromm-3]

I just took a look at the Amazon.com reviews for the Hawkins
book "On Intelligence". Most of the five-star reviews look faked,
and there are quite many of them. While his theory sounds a bit
superficial and oversimplified, he seems to be very good in
making buzz, promotion and advertising.

-J.

[Giles Bowkett]

It might have been cached by Google, you never know. Unfortunately the
podcast was basically all about business so there's zero scientific
content in it. (There is a little bit of stuff about why he set up the
institute which is kind of interesting.)

On 3/29/06, Jim Rutt <jim at jimrutt.com> wrote:

> I went out and spent an afternoon with Hawkins at his Redwood Nuerological
> Institute not long after the publication of _On Intelligence_ which I had
> found intriguing but frustrating in its lack of specificity.   At the time
> Hawkins was quite enamored of  (from Numenta web site)
>
> "Dileep George implemented a mathematical formalism of Hawkins' theory,
> demonstrating it is possible to express this new type of memory system in
> software"
>
> Indeed the MATLAB code was posted on the RNI website.  Alas, I got
> distracted on another project and by the time I got back to look at the
> code, it had been taken down and Hawkins had spun up Numenta.
>
> I wonder if the code is still available on one of the web archives?
>
> =jim rutt
>
>
> At 03:08 PM 3/22/2006, you wrote:
> >http://www.numenta.com/index.php
> >
> >Founder Jeff Hawkins also founded Palm, Handspring, and a neuroscience
> >institute which generated the research for this company. There's a
> >podcast of a speech he made on entrepreneurialism in general at
> >Stanford here:
> >
> >http://edcorner.stanford.edu/podcasting.shtml
> >
> >And somewhere a Real Audio thingy of an interview about his new
> >company which he did on/for NPR. (I have this idiosyncratic aversion
> >to RA so I haven't heard it, but the podcast is interesting.)
> >
> >
> >Giles
> >
> >============================================================
> >FRIAM Applied Complexity Group listserv
> >Meets Fridays 9a-11:30 at cafe at St. John's College
> >lectures, archives, unsubscribe, maps at http://www.friam.org
>
> ===================================
> Jim Rutt
> voice:  505-989-1115
>
>
>
> ============================================================
> FRIAM Applied Complexity Group listserv
> Meets Fridays 9a-11:30 at cafe at St. John's College
> lectures, archives, unsubscribe, maps at http://www.friam.org
>

[Jim Rutt]

turns out the MATLAB program is still arpund on the author's web site.

Here is a link to the paper
...http://www.stanford.edu/~dil/RNI/TechReport1/DilJeffNips2004.html

and here to various versions of the code:
http://www.stanford.edu/~dil/invariance/

At 02:35 PM 3/29/2006, you wrote:

>It might have been cached by Google, you never know. Unfortunately the
>podcast was basically all about business so there's zero scientific
>content in it. (There is a little bit of stuff about why he set up the
>institute which is kind of interesting.)
>
>On 3/29/06, Jim Rutt <jim at jimrutt.com> wrote:
> > I went out and spent an afternoon with Hawkins at his Redwood Nuerological
> > Institute not long after the publication of _On Intelligence_ which I had
> > found intriguing but frustrating in its lack of specificity.   At the time
> > Hawkins was quite enamored of  (from Numenta web site)
> >
> > "Dileep George implemented a mathematical formalism of Hawkins' theory,
> > demonstrating it is possible to express this new type of memory system in
> > software"
> >
> > Indeed the MATLAB code was posted on the RNI website.  Alas, I got
> > distracted on another project and by the time I got back to look at the
> > code, it had been taken down and Hawkins had spun up Numenta.
> >
> > I wonder if the code is still available on one of the web archives?
> >
> > =jim rutt
> >
> >
> > At 03:08 PM 3/22/2006, you wrote:
> > >http://www.numenta.com/index.php
> > >
> > >Founder Jeff Hawkins also founded Palm, Handspring, and a neuroscience
> > >institute which generated the research for this company. There's a
> > >podcast of a speech he made on entrepreneurialism in general at
> > >Stanford here:
> > >
> > >http://edcorner.stanford.edu/podcasting.shtml
> > >
> > >And somewhere a Real Audio thingy of an interview about his new
> > >company which he did on/for NPR. (I have this idiosyncratic aversion
> > >to RA so I haven't heard it, but the podcast is interesting.)
> > >
> > >
> > >Giles
> > >
> > >============================================================
> > >FRIAM Applied Complexity Group listserv
> > >Meets Fridays 9a-11:30 at cafe at St. John's College
> > >lectures, archives, unsubscribe, maps at http://www.friam.org
> >
> > ===================================
> > Jim Rutt
> > voice:  505-989-1115
> >
> >
> >
> > ============================================================
> > FRIAM Applied Complexity Group listserv
> > Meets Fridays 9a-11:30 at cafe at St. John's College
> > lectures, archives, unsubscribe, maps at http://www.friam.org
> >
>
>============================================================
>FRIAM Applied Complexity Group listserv
>Meets Fridays 9a-11:30 at cafe at St. John's College
>lectures, archives, unsubscribe, maps at http://www.friam.org

===================================
Jim Rutt
voice:  505-989-1115

[Steve Smith]

I was on "sabbatical" at Lawrence Berkeley Lab last year and did some
work with Don Glaser (Nobel Laureate for invention of the Bubble
Chamber) on his own neural models...  his team is working specifically
on the visual cortex which was partly how I got interested.

This lead me to become aware of Hawkins' work, Numenta and the Redwood
Institute.  I went to their open house and discovered that just like
his book, there are not a lot a of technical details left to the
reader's imagination.   Some of this is natural... intellectual
property and all.

I could not honestly evaluate what they were up to, but the optimist in
me is hopeful...

There is a local (Los Alamos) startup that roughly parallel's Numenta
but w/o the funding... but everything there is still hush-hush (I
think).  Anyone else here have wind of this effort?  Associative memory
model...

- Steve

On Mar 29, 2006, at 3:08 PM, Jim Rutt wrote:

> turns out the MATLAB program is still arpund on the author's web site.
>
> Here is a link to the paper
> ...http://www.stanford.edu/~dil/RNI/TechReport1/DilJeffNips2004.html
>
> and here to various versions of the code:
> http://www.stanford.edu/~dil/invariance/
>
> At 02:35 PM 3/29/2006, you wrote:
>> It might have been cached by Google, you never know. Unfortunately the
>> podcast was basically all about business so there's zero scientific
>> content in it. (There is a little bit of stuff about why he set up the
>> institute which is kind of interesting.)
>>
>> On 3/29/06, Jim Rutt <jim at jimrutt.com> wrote:
>>> I went out and spent an afternoon with Hawkins at his Redwood
>>> Nuerological
>>> Institute not long after the publication of _On Intelligence_ which
>>> I had
>>> found intriguing but frustrating in its lack of specificity.   At
>>> the time
>>> Hawkins was quite enamored of  (from Numenta web site)
>>>
>>> "Dileep George implemented a mathematical formalism of Hawkins'
>>> theory,
>>> demonstrating it is possible to express this new type of memory
>>> system in
>>> software"
>>>
>>> Indeed the MATLAB code was posted on the RNI website.  Alas, I got
>>> distracted on another project and by the time I got back to look at
>>> the
>>> code, it had been taken down and Hawkins had spun up Numenta.
>>>
>>> I wonder if the code is still available on one of the web archives?
>>>
>>> =jim rutt
>>>
>>>
>>> At 03:08 PM 3/22/2006, you wrote:
>>>> http://www.numenta.com/index.php
>>>>
>>>> Founder Jeff Hawkins also founded Palm, Handspring, and a
>>>> neuroscience
>>>> institute which generated the research for this company. There's a
>>>> podcast of a speech he made on entrepreneurialism in general at
>>>> Stanford here:
>>>>
>>>> http://edcorner.stanford.edu/podcasting.shtml
>>>>
>>>> And somewhere a Real Audio thingy of an interview about his new
>>>> company which he did on/for NPR. (I have this idiosyncratic aversion
>>>> to RA so I haven't heard it, but the podcast is interesting.)
>>>>
>>>>
>>>> Giles
>>>>
>>>> ============================================================
>>>> FRIAM Applied Complexity Group listserv
>>>> Meets Fridays 9a-11:30 at cafe at St. John's College
>>>> lectures, archives, unsubscribe, maps at http://www.friam.org
>>>
>>> ===================================
>>> Jim Rutt
>>> voice:  505-989-1115
>>>
>>>
>>>
>>> ============================================================
>>> FRIAM Applied Complexity Group listserv
>>> Meets Fridays 9a-11:30 at cafe at St. John's College
>>> lectures, archives, unsubscribe, maps at http://www.friam.org
>>>
>>
>> ============================================================
>> FRIAM Applied Complexity Group listserv
>> Meets Fridays 9a-11:30 at cafe at St. John's College
>> lectures, archives, unsubscribe, maps at http://www.friam.org
>
> ===================================
> Jim Rutt
> voice:  505-989-1115
>
>
>
> ============================================================
> FRIAM Applied Complexity Group listserv
> Meets Fridays 9a-11:30 at cafe at St. John's College
> lectures, archives, unsubscribe, maps at http://www.friam.org