John, (01)
>I changed the subject line to emphasize some issues that have been
>debated in AI for over half a century: Since digital computers are
>so radically different from the human brain, does it make any sense
>to imitate or even waste time by studying neural mechanisms? (02)
This is an interesting question and one that is also asked in the other
direction: are modern computers adequate models for neural computation? In
particular, there seems to be some debate about symbolic processing in the
brain. There, I find the work of CR Gallistel on prelinguistic thought in
bees and scrub jays interesting [1,2]. (03)
Aside from my professional work, as an artist I have a personal interest
in musical AI and on shifting away from the anthropocentric nature of
thought experiments and their apparent focus on natural language
interaction [3]. While it's exciting to think of interacting with a
human-like AI, I would like to see more work with regards to non-human AI
and what Gallistel calls "foundational abstractions" (e.g. numerosity)[4].
There is obviously a challenge here in how you communicate with the agent. (04)
With my current agent, there's a crude music-to-English controlled
proto-language. I currently communicate with it via music, but I hope to
shift it away from natural language more and more and to have it
communicate back with music and visualizations. This will make programming
a bit of a challenge in that I will have to replace the language in my
code with musical symbols or representations such as frequencies,
intervals, or other notation. At any rate, I guess my point is that
perhaps we should shift some attention away from modeling the human brain
to modeling the neural mechanisms of other species. (05)
>Moral: You can get a lot of insight from analyzing and reverse-
>engineering systems that successfully do what you're trying to do.
>But you don't have to imitate or simulate every detail. (06)
Cacioppo, Tassinary, and Berntson provide a fascinating history of
psychophysiology and it's shift from models informed by hydraulics and
mechanics (Galen c. 130-200), to electricity and magnetism (William
Gilberd 1544-1603), to machines (William Harvey, 1578-1657), to
electricity (Brazier, 1959). It is interesting to see the insights and
errors that each era's perspective engendered [5]. I believe we are now in
an era of biochemistry, imaging, and computation and we'll see where that
leads us. (07)
You may find these projects of interest: (08)
* http://www.cognitiveatlas.org/
* http://www.neuinfo.org/
* http://neurolex.org/wiki/Main_Page (09)
I presented something similar to Russell Poldrack's Cognitive Atlas work
at the Society for Neuroscience a few years ago. My work involved making
informal conceptual graphs of neuroscience concepts from a handful of
cognitive neuroscience textbooks tracing and relating the concepts and
putting the terms into a graph. I can pass on the poster and GraphML if
anybody is interested in that work. (010)
For cognitive architectures, I find Joaquin Fuster's work a good source of
inspiration [6]. (011)
Sincerely, (012)
Marcos (013)
[1] Gallistel, C. R. (2011) Prelinguistic thought.
<http://ruccs.rutgers.edu/~galliste/images/publications/Gallistel_2011_Prel
inguistic_Thought.pdf> in Language learning and development, 7, 253-262.
[2]
http://www.amazon.com/Memory-Computational-Brain-Cognitive-Neuroscience/dp/
1405122889
[3] http://0bservatory.org/mesomedes-intro/
[4] Gallistel, C. R. (2011). Mental magnitudes. Space, Time and Number in
the Brain: Searching for the foundations of mathematical thought.
<http://ruccs.rutgers.edu/~galliste/images/publications/Mental_Magnitudes.p
df>S. Dehaene and L. Brannon. New York, Elsevier: 3-12.
[5] Cacioppo, J. T., Tassinary, L. G., & Berntson, G. G. (2007). Handbook
of psychophysiology, 3rd edition. New York: Cambridge University Press.
First chapter of 2nd edition is available through Library of Congress:
http://catdir.loc.gov/catdir/samples/cam032/99036718.pdf (014)
[6] J.M. Fuster - The prefrontal cortex-An update: Time is of the essence.
Neuron, 2:319-333, 2001. (015)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J (016)
|