Software Engineering discussion

6 views
Gödel, Escher, Bach > ... Ant Fugue

Comments Showing 1-2 of 2 (2 new)    post a comment »
dateUp arrow    newest »

message 1: by Erik (last edited Oct 19, 2010 11:11AM) (new)

Erik | 165 comments This dialog starts off with some references to reductionism versus holism, which seems like it wraps up the previous chapter's concepts. When the dialog turns to ants and behaviors of ant colonies, they seem to be introducing the "emerging properties" concepts. I thought I was getting some foreshadow in the previous chapter about these concepts, but I wasn't sure until reading this dialog. There are several areas in the MIT videos that discuss emerging properties, which is the only reason I knew to look for it.

I thought the "map your brain onto an ant brain" line was very funny.

This was a nice dialog to follow the previous "software" chapter. I was still thinking about modeling the Ant and behaviors with software threading and distributed work ideas. I also like thinking about the "Traveling Salesman" problem, which can be attacked with an Ant-like behavior approach.

The "group will stick together if its size exceeds a threshold" is equivalent to "a group will stick together if it has something to do" statement caught my attention. I'm not sure I understand the equivalency, but it does remind me of start-up-company-like behavior.

This dialog has a lot of really great word selections. The "emergence of various mechanisms... responses to external pressures" was great with the recent discussion of gasses and pressure.
I recently heard someone quote Einstein as saying something like "if you can't explain it to a 6 year old, you don't understand it". I was the 6 year old reading this chapter - outstanding dialog (advanced concepts presented in an understandable way).

I was reading this chapter while on an airplane. As my plane was landing, I smiled at the irony of watching the cars and people looking like ants from up in the air.


message 2: by [deleted user] (new)

I am just returning from a few conferences, and one clear trend is the growing interest in harnessing large number of processes that know only about local behavior to support a global, emergent behavior. As the growth in individual processor speed slows, and multiple core horizontal growth becomes more and more predominate, the interest in these algorithms will continue to grow.

Here is one example:
http://portal.acm.org/citation.cfm?id...


back to top