Comparison of Cybernetic Insect to Real Insect (Beetle)

I thought I would start with a simple comparison showing the various parts of a real insect (a beetle) and how they compare with the various parts of the Cybernetic Insects at their base level (simplest circuitry).  Both exhibit emergent behavior, both have rudimentary sensing skills, both remember and gain skills as they experience their world, both are nocturnal, and both need to feed to gain energy (beetle – real food, cybernetic insect – light).  Both do not like light, but must come into the light at times to feed.  In the case of the cybernetic insect – light is the only thing it feeds on.

Both contain rudimentary neural networks.  In the beetle’s case it is a simple biological neural network.  In the case of the cybernetic insect, it is a simplistic analog neural network (i.e., made with discrete components).

CyberInsect Drawing - Physiology

Both have basic instincts.  In the cybernetic insect’s case this is facilitated with functions that it must satisfy. Both must learn to interface with their world.  The cybernetic insect must figure out how to walk, how to take input in from its sensors and plan its movements; whether it’s looking for darkness or reacting to running into an obstacle.

 

The Mad Scientist “Branches Out”

The life of a Mad Scientist can be very stressful… you know… all those Villagers with their pitchforks and torches invading your privacy all the time.

Angry VillagersSo occasionally you have to go somewhere, grab a really good cup of coffee, and write – in my case it comes out as poetry.   Here are some links to my poetry.  Next time I promise to tell you more about the original bugs and include some videos.

1. Life’s Memories:  http://www.amazon.com/Lifes-Memories-James-Crowder-ebook/dp/B008ZDRIZE

2. Solomon’s Darkness:  http://www.amazon.com/Solomons-Darkness-James-Crowder-ebook/dp/B008SXAQ78

3. Being Human:  http://www.amazon.com/Being-Human-James-Crowder-ebook/dp/B00BPXHCOS

4. Unicorns Step Lightly:  http://www.amazon.com/Unicorns-Step-Lightly-James-Crowder-ebook/dp/B005OU65S2

5. Fragile as an Icicle:  http://www.amazon.com/Fragile-as-Icicle-James-Crowder-ebook/dp/B005P4I1K2

 

 

Papers and Books on Systems Engineering and Artificial Intelligence

Tags

, ,

I thought I would provide some links to a few of the papers and books that I, my daughter Shelli Friess, and my colleague Dr. John Carbone, have published over the last few years to give you a flavor of our research into Artificial Intelligence and Artificial Cognition.

1. Artificial Cognition Architectures, Springer Publishing, 2013: http://www.amazon.com/Artificial-Cognition-Architectures-James-Crowder/dp/146148071X

2. Systems Engineering Agile Design Methodologies, Springer Publishing, 2013: http://www.springer.com/engineering/mechanical+engineering/book/978-1-4614-6662-8

3. Occam Learning through Pattern Discovery: Computational Mechanics in AI Systems: http://worldcomp-proceedings.com/proc/p2011/ICA3797.pdf

4. Autonomous Creation and Detection of Procedural Memory Scripts: http://worldcomp-proceedings.com/proc/p2012/ICA4256.pdf

5. Applications for Intelligent Information Agents (I2As): Learning Agents for Autonomous Space Asset Management (LAASAM): http://worldcomp-proceedings.com/proc/p2012/ICA2059.pdf

6. Recombinant Knowledge Relativity Threads for Contextual Knowledge Storage: http://worldcomp-proceedings.com/proc/p2011/ICA2772.pdf

7. Knowledge Density Mapping for Derivation of Inference Potential: http://worldcomp-proceedings.com/proc/p2011/ICA2773.pdf

8. Flexible Object Architectures for Hybrid Neural Processing Systems: http://enu.kz/repository/2010/AIAA-2010-3519.pdf

 

And just please remember, Skynet was a Hollywood fabrication – :-):

1. http://connection.ebscohost.com/c/articles/3592170/bt-bae-lockheed-martin-team-skynet-5-bid

2. http://skynetisreal.com/tag/lockheed-martin/

 

The Cybernetic Insects – History

Tags

, , ,

About 15 years ago I began a serious study and research of Artificial Intelligence.  Early on I began to look at analog versions of Artificial Neural Networks as well as digital versions.  The challenge was whether I could create a stable analog neuron.  The result was the 1st generation cybernetic insects.  They each had one neuron, a motor controller, and a timer chip that is used to regulate current flow through the analog neural structure.  I discovered that subtle changes in the resistance/capacitance networks around the transistor that forms the neuron affect the shape of the leading edge of voltage pulses and affect learning within the analog neural framework.  The trick was to design a circuit that was stable and repeatable without tuning.

Even with their simplistic analog neural structure, the cybernetic insects learned (which in this case amounts to remembering and skill retention) and behaved entirely differently from each other, even though they started out roughly the same (same components, but each component will be subtly different).  Again, subtle differences in the feedback impedance between the base of the synapse and the external poditic terminal of the neuron make a difference in the positive feedback and thus, affects their learning.

Over time I will be telling the story of their evolution, including videos of the original artificial life forms… who at the time did not have names as I hadn’t completely anthropomorphized them yet.  Taking a chapter from Dr. Seuss – we will call them Insect 1 and Insect 2.  Then you will meet the 2nd generation: Zeus, Hercules, and Athena.  I hope you’ll find this educational and entertaining.  As the year progresses, I will be designing, building, and testing the 3rd generation artificially intelligent, cybernetic, artificial life forms – getting ready for next year’s artificial intelligence conferences.