The working title for my thesis has been “Empirical Prediction as the Stuff of Knowledge” for quite some time. I have been fond of this title, though not totally sold on it. Rich has not been fond of it, particularly “stuff of.” I thought it was because it was too informal, but I think now he found it too vague?
Anyway, the new working title is “An Empirical Approach to Knowledge Representation.” This is better. Changing from empirical prediction as a noun (which was a jargon-alert) to empirical as an adjective helps with accessibility. And then good old “An” is always nice. Not claiming to be the definitive work!
I was resisting “knowledge representation” for a while, because I thought the term had been lost to the old-school approaches. But now that I know what philosophical framework we fall into, I’m more comfortable with it. The agent is using a representation, and the thing being represented is patterns in the sensorimotor data rather than external entities.
It was a great meeting on Friday, and I have an improved working title: simple, clear, accurate. Good things.
Leave a Reply