a particular interest of mine is, and has been for sometime, about
finding patterns in things; working out where the logical train of
inference goes, in order to be able to tell a computer about it;
simplifiying things to their smallest elements. Things I have enjoyed
doing:

learning new programming languages: not relevant. This with my
rather experimental learning technique. a) can feel I'm getting
somewhere. b)can feel I'm working against something. not sure really.

sitemap code: over and over, trying to work out what logical process
*I* use to make a sitemap, and teach the program that.

C code: new language, challenge of wanting to be able to do it.

both sitemap and dove: things that I thought I could finish quite
quickly, and that turned out to drag on; each day was "just let me
finish this" and would end up with me 'about' to get dressed/eat
several hours later. 

1B project: particularly interested in as much automation as possible,
similarly guild website. simplifying things so that management could
be done without human interference. cf other tester who hard coded
more things in.

part II project. bored! language issues; just look stuff up. stupid
irritating memory bugs. lack of interest in the technique and the end
result.

ringing: patterns patterns patterns. once you've "got" the point of
a method, it reduces from a long squiggly line to some fairly simple
rule. and then there are more and more patterns that one can spot,
deduce, etc. 


Eng lit: analysis. reading books; ok fun. learnt at alevel to search
for the point, the meaning, the "shape" of a poem. shape of things,
good. simplification again. like to see historical events in shapes
and colours too. (that one memorable essay in which I expressed
that). looking for a general picture, for a bigger picture, for
extracting a meaning from the text, patterns again? similar thoughts
about livejournal.

understanding proofs. I don't like constructing proofs. I'm not
particularly happy proving things. I do like understanding proofs; in
some respects, the same.. simplification thing.  Once you've got the
point of the proof, the lines i hangs on, the rest should follow, if
it's a nice proof...
 
so...  patterns patterns patterns. looking for patterns, looking
for ways of (automatically) extracting information from data, thinking
about whether computers can do this where humans can't, teaching
computers to look for and find patterns, finding patterns oneself in
order to teach them to computers. "I can do something like infer a
sitemap from a website, or sort spam from nonspam. What rules do I
use? How much *can* I teach the computer? Could someone less
intelligent than me do this? Then can a program?" What about if I give
it lots of websites and sitemaps? Can it infer how to do it? how will
it know what kind of things to look for?