I'll let you know in detail this week-end : I'll have to fetch his thesis
at my place.
In fact his work entailed the construction of a generic A-Life simulation
tool in C++.
The intro to his thesis is quite well done : he starts practically from
scratch to first justify the use of the computer (inspired from Joel de
Rosnay's "Le Macroscope").
Renaud (that's his name) finally did a simulation of sheep in a grassy
environment. Each sheep has a simple neural network and there's a
genetic algorithm working in the background.
Very nice work.
I was just talking to him and asked him if he knew about A-Life (because
here in France, they're still lagging behind in terms of algorithms
and such and GA, and ALife are not yet popular even in the computer science
circles) and he answered casually "yep done a thesis on that.".
That was the start of a series of interesting discussions. Except he hasn't
read much about evolution and he isn't a good practitioner of the systemic
approach and he did not know about Turchin's Meta-System Transition theory,
and hence the organ-organism-population metaphor cannot be shifted upward
of the human being for him (just like D.S. Wilson in his paper).
He likes Physics as well and maths, and so there's room for many varied
and intertwining interesting issues to be discussed.
Yash.