|
|
Once upon a time, there was an operating system called Windows 3.1. In the office where I worked, we used programs running under this platform, and regularly had problems of the "doesn't print" kind. Though I grew
quite good at fixing them, I couldn't figure out their cause and I eventually came to the conclusion that the "let's try this and see if it works" approach just wasn't good enough. What I needed was some formal education which would
teach me how to solve all computer problems once and for all.
Looking back, I am grateful there were so many bugs in Windows, because if I had not been fed up with the inefficiency of empirical troubleshooting I might never have decided to study computer science. At the start of my first year
in Bath University, I asked my course director why we didn't have a troubleshooting course, and he laughed. He told me not to worry, that I would learn how to do this, that I would have no choice in the matter. This wasn't what I
wanted to hear, but by that time I was already finding my studies very interesting, so my original goal didn't really matter to me anymore.
We were four women in my class, for about thirty male students. I found this discrepancy mildly puzzling, but to understand it I would have had to ask people who /were not/
there, so I let it go. I never felt discriminated against, nor was I favoured because of my gender. We four women banded together. We always did group coursework together and we sometimes, illegally, did individual coursework
together. Even when we worked out the exercises on our own, we compared our results before handing in our worksheets, and when our results didn't match we worked out why. At the end of the three years, nobody in my class go a first
degree. Out of the five people who got an upper second, three were women, myself among them.
The best teacher I ever had was Dan Richardson. I was very priviledged to have him as a lecturer during all of my three years in Bath. He would often start a lecture by writing a theorem or a mathematical formula on the board (he
seldom used slides). Then he would turn round and look at us. We dutifully read the theorem or formula, but sometimes we didn't undertand any of it. When Dan saw a lot of blank faces staring hopelessly at the board, he would smile
very slightly and make a small dismissive gesture with his hand. "Don't worry, he would tell us, you'll understand". He was always true to his word. By the end of the three years, I trusted him completely. He could start a lecture
by filling the board with abstruse mathematics, if he promised us we would understand I knew I just had to sit back and listen and by the end of the lecture it would all be clear in my mind. He was that good.
If any single person has to bear the responsability for my choice of career, it is Dan Richarson. When he taught us his "Introduction to AI" lectures they were so facinating that I became overwelmed by the urge to learn more about
it. Though at the time I knew about Data, the android lieutenant in the starship Enterprise of the Star Trek "Next Generation" series, I had never before considered the possibility that there could actually be such things as
artificial neural networks. Hearing about them for the first time, and, thanks to Dan, undertanding how they worked, was a big shock. Though Dan told us about their limitations, I immediately imagined a simulated human brain running
on a pc, and I had been a science fiction fan for long enough to have many wild ideas about what this could lead to.
Once I had my bachelor degree, I went to Bristol University to do an MSc in Machine Learning. There, again, I was one of a minority of women, but the ratio was up to about a third. There were also women among the faculty, Nada
Lavrac among them. Nada gave us lectures about the intellectually challenging workings of ILP/first order machine learning algorithms, and we all knew she was a recognised researcher in the field. Had I felt the need for a role
model I had one of the very best right in front of me. As it were, I liked and respected her, but I never identified with her.
After Bristol, I went to Lyon 2 University to do a DEA. There, several groups of students had intermingling timetables, so depending upon the lecture there were sometimes more women and sometimes more men. An inspiring lecturer was
Michele Sebag, who radiated so much enthusiasm about genetic algorithms that it was quite contagious. I enjoyed her lectures and her friendly attitude, but I didn't identify with her either. I am now a PhD student working under the
supervision of Yves Kodratoff. I still don't have a female role model, nor, for that matter, do I have a male one. I guess I am just not the type of person who needs that kind of relationship to stay motivated. What motivates me is
a deep curiosity concerning my field of work and AI in general. I owe this curiosity to Dan, and I hope it never becomes satiated.
|
