[ULS-SIG] Melanie Mitchell

Duncan McGreggor duncan.mcgreggor at gmail.com
Mon Dec 14 22:01:41 CET 2009


Wow, with seconds of sending this, a friend of mine (who's not
subscribed to this list) sent me the following new item:
  http://web.mit.edu/newsoffice/2009/ai-overview.html

At fist glance, it doesn't look like there's much intersection between
the two topics, but then I read these quotes:

Instead, he says, what’s needed are ways to “make systems made up of
lots of pieces” that work together like the different elements of the
mind. “Instead of searching for silver bullets, we’re looking at a
range of models, trying to integrate them and aggregate them,” he
says.

and

A new approach to programming, called RALA (for reconfigurable
asynchronous logic automata) attempts to “re-implement all of computer
science on a base that looks like physics,” he says, representing
computations “in a way that has physical units of time and space, so
the description of the system aligns with the system it represents.”
This could lead to making computers that “run with the fine-grained
parallelism the brain uses,” he says.

Sounds pretty far out there. Even further than ULS systems! But not so
dissimilar ;-)

d

On Mon, Dec 14, 2009 at 1:52 PM, Duncan McGreggor
<duncan.mcgreggor at gmail.com> wrote:
> I've just discovered a brilliant mind in fields that are of great
> interest to ULS researchers: Melanie Mitchell. She's conducted
> research and has written extensively on the topics of genetic
> programming, cellular automata, and computational perception. I'm
> currently reading her paper entitled "Evolving globally synchronized
> cellular automata" and it's quite fascinating. Here's a great quote
> from the introduction:
>
> "Given the widespread appearance of synchronization in decentralized
> and spatially ex-
> tended systems in nature, evidently evolution has successfully
> overcome this problem. Evo-
> lution has effectively taken advantage of the spatially local dynamics
> in its production of
> organisms which, on the one hand, consist of potentially independent
> subsystems, but whose
> behavior and survival, on the other hand, rely on emergent
> synchronization. These observa-
> tions leave us with an unanswered but biologically significant
> question. By what mechanisms
> does evolution take advantage of nature’s inherent dynamics?"
>
> She concludes an overview of their work in section 4:
>
> "...So we are faced with a problem familiar to biologists and
> increasingly familiar to
> evolutionary computationalists: how do we understand the successful
> complex systems (e.g.,
> φ) that our GA has constructed?"
>
> She later describes the mechanism they employed to analyze their
> results, and that bit was *very* fascinating! In previous work, one of
> her frequent collaborators helped develop a theoretical abstraction, a
> means of explaining complex systems. The mechanism they propose and
> that has helped them analyze these results (as well as previous ones)
> involves looking at the produced systems in terms of domains,
> particles, and particle interactions. (J. P. Crutchfield and J. E.
> Hanson. (1993). Turbulent pattern bases for cellular automata. Physica
> D, 69:279–301.)
>
> Here's her list of published papers:
>  http://web.cecs.pdx.edu/~mm/publications.html
>
> The reason I'm looking into this (beyond the fact that such research
> has been identified as crucial for ULS systems) is that the
> "self-organizing object meshes" work (play) that I'm hoping to get
> back to soon can be cast in terms of cellular automata. I fully
> anticipate that these papers and the other writings of Dr. Mitchell
> will provide endless inspiration for exploring implementation avenues
> :-)
>
> Anyway, I will continue to consume this and related research with much joy!
>
> If anyone else gets around to reading her stuff, I'd love to hear what
> you think about it...
>
> d
>
> P.S. For those of you that may remember the book "Gödel, Escher, and
> Bach", Douglas Hofstadter was one of Dr. Mitchel's doctoral advisers
> at the University of Michigan :-)
>


More information about the ULS-SIG mailing list