Jeff, Is the workshop on Sept. 18 by invitation only?
-Steve O. On Sep 5, 2013, at 6:23 AM, "Jeff Hawkins" <[email protected]> wrote: > Dan Hammerstrom at DARPA is working at putting together a DARPA program > called the Cortical Processor. So far there was a 1.5 day workshop earlier > this year in Wash D.C. and there is another one scheduled for September 18 > here in the bay area. I spoke at the earlier one and will be speaking at the > upcoming one too. The goal of the program is to create new HW architectures > designed specifically to model the neocortex. The two workshops are to help > DARPA gather ideas for the program. If all goes well DARPA will formally > launch the program around the beginning of the new year. DARPA does not do > the actual work, they provide funding for teams that compete to meet a set of > goals. Usually these programs run for several years. > > Dan sees the CLA as a prime example of the kind of algorithm he would like > the Cortical Processor to implement. Dan has a long history building neural > network type chips and he understands the CLA deeply, so he knows what he is > doing. The NuPIC community can play an important role in this. The fact > that the CLA is open sourced helps, the fact that there are people studying > the CLA helps, and of course any commercial applications of the CLA helps. > > As I have said before today is similar to the 1940s in the computing era. > Back then they were just starting to understand how to build computers. They > had not yet invented operating systems, compilers, transistors, disk drives, > and so on, but they had the basic idea of a data store that contained both > instructions and data and that this could compute anything. Once the basics > were understood, progress was steady. This is where we are today in building > intelligent machines. We understand the basics of hierarchy, SDRs, sequence > memory, etc. and we are trying to figure out the best way forward. BTW > government funding played an important role in the creation of the first > computers and I believe Dan Hammerstrom believes DARPA can play a similar > role again. > Jeff > > From: nupic [mailto:[email protected]] On Behalf Of Matthew > Taylor > Sent: Wednesday, September 04, 2013 9:02 PM > To: NuPIC general mailing list. > Subject: Re: [nupic-dev] DARPA RFI > > Jeff wrote a blog post about this, too: > http://numenta.org/blog/2013/08/13/brains-and-machine-intelligence-a-long-time-coming.html > > --------- > Matt Taylor > OS Community Flag-Bearer > Numenta > > > On Wed, Sep 4, 2013 at 8:49 PM, Alexander van Dijk <[email protected]> > wrote: > Just ran across this article > > http://www.networkworld.com/community/blog/darpa-wants-computers-fuse-higher-human-brain-function > > which mentions the release (August 14) of an RFI by DARPA asking for "Request > for Information (RFI) on Research and Development of a Cortical Processor" > > https://www.fbo.gov/index?s=opportunity&mode=form&id=91bc9e58d6fa024d55d7c0583d38fc21&tab=core&_cview=0 > > I couldn't find the same RFI on the DARPA website, but Daniel Hammerstrom is > mentioned as the primary POC > > http://www.darpa.mil/Our_Work/MTO/Personnel/ > > Alexander > > _______________________________________________ > nupic mailing list > [email protected] > http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org > > > _______________________________________________ > nupic mailing list > [email protected] > http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org
_______________________________________________ nupic mailing list [email protected] http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org
