What's this all about?

Having had an interest in Artificial Intelligence for over 15 years I finally moved from a software engineering role in 2016 to a role where I actually get to design and implement Machine Learning based applications. As far as I recall, my interest in AI began when I took a course on Social and Political Perspectives on Information Technology as part of a Masters in Business Studies in Dublin back in 2001. That’s when I first heard of Hubert Dreyfus’ book What Computers Still Can’t Do, whose title intrigued me no end. Also around then I read Douglas Hofstadter’s Gödel, Escher, Bach for the first time. I’ve read it another couple of times since - it is quite possibly my favourite book ever :)

My primary interest is in Natural Language Processing. I’ve always had a love of language, having studied German and French at university. My current work involves making use of Word Embeddings, a technique that I became fascinated by from the very first moment I heard about it. Although I have an accute sense of the enormous difficulty faced by anyone trying to get computers to understand language (without common sense knowledge, this is quite simply an impossible task), I see the idea of learning distributed representations from millions of examples of word usage as extremely elegant and obviously useful. It’s not about passing the Turing test, it’s about pragmatic applications involving raw text, e.g. content classification.

For those interested, this blog was built using Hugo, which I must say is a joy to use.