Happy new year! I began teaching my seniors Artificial Intelligence theory today. this is my fifth year teaching it and each year I look forward to it more and more. I don’t claim to be anything resembling an expert but really enjoy the discussions and tangents that the topic generates.
today my lesson focussed on definitions of intelligence and how AI is defined. I’d argue that AI is never going to make progress so long as the research teams concentrate on the logical or constructivist areas – this to me seems a very male-oriented idea based on programming fundamentals – instead focussing on the emotional element and making this work should play a big part in driving AI research forward. It makes sense: we learn because we see something in it for ourselves, whether that is money, prospects, praise or enjoyment we are driven to learn by our feelings. In the past AI research has concentrated on mimicking conversation, vision, categorisation of objects or information but the machine has no ulterior motive to improve. Young babies recognise the faces of their parents or siblings because they have an emotional attachment – love is pivotal to their development. Toddlers learn to walk because standing tall increases the area they can explore and moving while standing is faster than shuffling while sitting – curiosity fires the learning. Surely if the researchers focus on getting emotional simulation right first the other areas would make leaps and bounds?
This is an open discussion that I share with my senior class – all comments are welcome especially disagreements!