Finishing up outstanding action items, mainly getting docs and a repo setup.

[Noah Gift]

To keep myself motivated on some of the things we are doing on this
list, I am going to commit to publishing a weekly update of what we
discuss on the list on my blog, which then goes to Planet Python.
Here is the summary for this Sunday:
http://artificialcode.blogspot.com/2009/04/python-artificial-intelligence-sig.html

The biggest outstanding action item seems to be a domain,
documentation and a repository. I personally like Jeff Rush's idea
of using ai.python.org and Sphinx, because it solves the whole problem
of wiki permissions, etc, it becomes a simple mercurial repository
where code and docs are shared. Is it ok to go forward with this
Jeff?

Once this gets put in place then it might be nice to start breaking
down docs into categories such as pre-baked solutions, projects,
how-to's etc. The other nice thing about using Sphinx is that we have
a nightly toolchain that builds a PDF.

For the repository we could use bitbucket, or google code. Bitbucket
seems like a reasonable choice for now because it supports openid,
although maybe google code's hg support will be awesome.

--
Cheers,

Noah

[Tennessee Leeuwenburg]

I think that is a fantastic idea -- it will definitely help keep interest in the group if you are happy to keep compiling these status updates. I think getting a wiki going will be important for our collaboration.

I've been spending some time lately looking for more resources in the areas of AI in which I am most interested. I think it would be of great value to list each interest person on the wiki together with a list of topics and/or papers which talk about their areas of interest. There is obviously a fairly active research community in many/most areas of AI which I think we will need to be conscious of in directing any group activities.

I'd certainly be interested in putting together such a page for myself to help explain what my AI interests are and also as a kind of collection point for things I find interesting on the topic.

I'd also be interested in talking to anyone who is interested in overall AI architecture design. I have some ideas on this topic which I am very slowly attempting to turn into a primitive prototype system. I'd love to put together a basic AI architecture, then provide implementation samples of a lot of standard AI techniques to hang them all together and provide a system anyone could plug into to play and experiment with ideas. However, I really don't have the time to put in to actually get something of that scale right. I'd really like to connect with interested people though to just talk about the ideas in a fairly freeform way, then let whatever come what may.

Cheers,

-T

--
--------------------------------------------------
Tennessee Leeuwenburg
http://myownhat.blogspot.com/
"Don't believe everything you think"