2014 Annual International Conference on Biologically Inspired Cognitive Architectures (BICA 2014)
Fifth Annual Meeting of the BICA Society
November 7-9 (Friday-Sunday): Massachusetts Institute of Technology, Cambridge,MA 02139, USA (http://bicasociety.org/meetings/2014).
Sponsors: The BICA Society; The MIT; Elsevier B.V.; DOLLabs, Inc. Points of contact: Paul Robertson (
paulr@dollabs.com) and Alexei Samsonovich (alexei@bicasociety.org)
 

Mutual Interest - Industry Meeting (MI-IM) Session


Introduction

The purpose of the MI-IM track is to help people from Academia, Industry, and other groups to find each other and establish partnership and collaborations. To prepare for it, we conduct a video-survey following the format of Videopanels (http://bicasociety.org/vp/archives/). The MI-IM Track is scheduled for the Saturday afternoon (please see the conference program) and will include a sequence of short presentations in parallel with small-group discussions. All BICA 2014 Participants are encouraged to express their interest and participate in MI-IM. Please start by sending an email to Alexei Samsonovich.

General Questions

  • What do you expect from this BICA conference, or would like to see in it?
  • What funded projects are you pursuing or planning to pursue that others should be aware of?
  • What missing capabilities and in what time frame do you need for your projects?
  • What technologies / capabilities can you provide to collaborators?
  • What new funded research program would you like to see created in the near future?
  • Any additional comments or open questions that people should be aware of?

First Answers

A Videopanel with Lee Scheffler and Alex James

[ Click below - or watch it on YouTube at http://youtu.be/yqIcB0Z7keU ]



On Oct 18, 2014, at 7:42 PM, Lee Scheffler wrote:
Here are my answers to the questions below:
- What do you expect from this BICA conference, or would like to see in it?
a) Keep current on the many research and development fronts
b) Introduce a new technology (NeurOS and NeuroBlocks) to accelerate cognitive research and development, and hopefully kindle an interest community
- What funded projects are you pursuing or planning to pursue that others should be aware of?
a) The work is currently self-funded (and seeking support/funding), with no current industry, academic or government affiliations.
b) I am planning to pursue the many potential vectors listed below as funding and collaboration allows.
- What missing capabilities and in what time frame do you need for your projects?
a) The technology is purposely open, to enable easy integration of/with/in other existing technologies with growing synergy, and to allow it to be incorporated in other research and development programs, hopefully accelerating them.
b) There are many vectors to grow and take advantage of this capability: new/improved components, incorporating other existing technologies, experimenting with constructs for multiple cognitive challenges, porting to multiple platforms, tool enhancements, etc.
c) Building working models of cognitive functions based on biologically plausible components and interconnections may be a route to understanding biological constructs.
d) I hope to kindle a growing broad-based collaborative effort at whatever pace makes sense.
- What technologies / capabilities can you provide to collaborators?
a) By the conference date, I expect to have a web site with descriptive material and videos showing demonstrations of use of the technology on a variety of cognitive tasks. This will grow steadily, hopefully helping interested others gain familiarity with the technology and its potential.
b) The web site will initially offer an email signup interest list and hopefully grow into an open forum about using the technology to advance cognitive R&D.
c) I am hoping to have a development toolkit available for distribution within 6 months, including the easy ability to integrate other technologies.
d) The technology supports creation of redistributable components, sub-assemblies and whole applications. If NeurOS usage grows, I expect to add a component/sub-assembly/application exchange to encourage community synergy.
e) I am open to other ideas to spread awareness. And if it proves as interesting as my early responses have suggested, to encourage collaboration.
- What new funded research program would you like to see created in the near future?
a) I would love to attract funding on any and all potential vectors of the technology.
- Any additional comments or open questions that people should be aware of?
• As you can see from the above, BICA 2014 is my first planting of a seed of something I think (in my humble opinion!) has great potential.
• I chose BICA for this because it's openness, breadth and biological grounding seemed like a good place to start and a good fit for my thinking. The BICA comparative architectures chart was very helpful in me getting my initial bearings in this research and technology space when I began my research.

--Lee Scheffler


On Oct 19, 2014, at 9:38 AM, Alex James wrote:
Industry meeting would be an excellent idea - especially to develop industry linked projects.

Scaling and implementation issues of on chip neural circuits - would be a topic of discussion following the BioTL symposium - can take a lead on this.

- What do you expect from this BICA conference, or would like to see in it?
To build new collaborations. In the multidisciplinary areas of neuroscience, computing and circuits.

- What funded projects are you pursuing or planning to pursue that others should be aware of?
Currently, involved in funded project on bioispired neural chip design and few projects awaiting funding.

- What missing capabilities and in what time frame do you need for your projects?
We need more expertise in neuron sciences and networks.

- What technologies / capabilities can you provide to collaborators?
IC design, device characterization, Memristor circuits, CMOS-hybrid arrays

- What new funded research program would you like to see created in the near future?
Projects in the area of Brain inspired circuits and systems.

- Any additional comments or open questions that people should be aware of?
I would be happy to accept applications from post-docs and faculty to join our department at Nazarbayev University.

Cheers
Alex


A response from L. Andrew Coward


Hi Alexei,

In response to the questions, at ANU we are pursuing a project that would focus on implementing an electronic model that mapped closely to physiology.

Perhaps the key advantages of the brain over current electronic learning systems are 1. the ability to learn many different functions; 2. the ability to learn with very limited degradation of past learning; and 3. the ability to learn from single experiences.

A theoretical model for how the brain achieves this has been proposed, down to neuron information processes. The concept is to iterate between implementations of the model and neuroscience experiments to test how the model matches anatomy and to get ideas for improving the model. The neuroscience experiments would investigate neuron connectivity and integration algorithms, but in a targetted fashion guided by model predictions rather than just accumulating huge amounts of data.

There is a proposed program by IARPA which would be a good fit for our proposal. One thing we are missing is a company that is interested in implementing a system that goes beyond the more prototype systems that can be implemented in an academic environment.

A panel for the video would include a couple of people from the Neuroscience school at ANU, but I am not on site at the moment and it would be hard to arrange on your timescale. So maybe just use this written response.

Best regards

Andrew


Advanced Web Counter