Grey Thumb Silicon Valley

Next Meeting

Note: Got an A-life system? Bring it to show/talk about!

Note2: Per Grey Thumb traditions we may retire to a local establishment following the meeting

1) Synthetic biology: Why it is the next best thing since sliced bread and how we'll be settling the solar system because of it.
John Cumbers, Graduate Student, NASA Ames Research Center & Brown University.

You cannot escape the press, investment, hype and excitement about synthetic biology. And rightly so, this new field is taking genetic engineering and dragging it kicking and screaming into the 21st century. Whilst the last 30 years of recombinant DNA technology has produced major advances in our understanding and ability, the design and construction of biological systems still remains an ad hoc process, where solutions are custom designed for a specific problem or task and little is generalizable. This is where synthetic biology comes in, with the thinking that if we invest in foundational research then the engineering of biology will become much easier.

I'll first give an interactive overview of the biology covered in the talk to make sure that we don't lose anyone along the way. I'll show examples of foundational synthetic biology and then talk about what I hope we'll be doing with photosynthetic extremophiles once the cost of synthesizing whole genomes has fallen to just a few thousand dollars.

Read more about John and his work at:
http://www.cumbers.com/

2) Jeffrey Ventrella Update
Jeffrey Ventrella, Internet Archive

The first half of the presentation is the Musical Gene Pool (published as a chapter in this book: http://www.springerlink.com/content/k00g03x142k153l9/ ). The Java applet is here: http://www.ventrella.com/EvoMusic/EvoMusic.html Hundreds if MIDI events are encouraged by a listener/user to link up into phrases, while the listener/user listens and clicks when something sounds cool . After being rewarded, the MIDI events acquire fitness values and perform the various genetic algorithm acts we all know and love.

The second presentation about little things that make something big is "Gliders and Riders" - where a grid of 2D cellular automata with initially heterogeneous rules is gradually made homogeneous by way of a particle swarm that rides along the chaotic waves of the CA and encourages coherent spacetime patterns to emerge (gliders). The particles perform the GA acts we all know and love on the CA rules whenever they enjoy a good ride. This work was published as a chapter in this book: http://www.springerlink.com/content/25u765j72j431714/ The Java applet is here... http://www.ventrella.com/Alife/Cells/GlidersAndRiders/Cells.html

3) Bruce Damer, Brief Update
Bruce Damer, Biota.org

Bruce will give a brief update on Biota.org, the podcast, his upcoming trip to London and the Grey Thumb meeting there, and the EvoGrid concept development.

www.archive.org
116 Sheridan Ave
San Francisco, CA 94129
(415) 561-6767

Previous Meeting

Photos

Attendees: 1 2 3 4 5
Macropond: 1 2 3 4
Maker Faire Robot: 1 2 3 4 5
Gene Pool: 1
Lunar Excavator: 1

Abstract

Macropond is a simple and eventually open source artificial life simulation, as well as a potentially useful testbed for evolvable instruction sets.

Each Macropond environment resides in its own thread of execution. This makes it very easy to create multiple Macroworld environments that run simultaneously, which is useful for comparing the effects of
minor changes, or even in comparing two different instruction sets, while also maximizing performance on dual core processors.

Critters in Macropond are one to sixteen segments long (with each segment displaying as a pixel), and each may have a much longer instruction set. They have a head, which can be oriented in four directions, and they move as worms with their body following the head. The basic instruction and segment type is Photosynthesize, which gains energy towards reproducing and displays as a green dot.

However, with reproduction comes a chance of a random mutation, and when a critter with a Move instruction moves onto a green dot it will eat the critter it belongs to. This competition drives the evolution of various types of critters - tending to resolve roughly into "plants" and "animals", although since the "plants" tend to move to avoid getting eaten, it's more reminiscent of a predator-prey relationship.

Unlike most other CA-like simulations, Macropond has a nice zoom view that shows the critters as discreet organisms along with their energy levels and head orientation. This has been a great help in insuring that the simulation is functioning as intended.