Sometimes one statement can change your behaviour

One statement can change change your behaviour

The expert system dogma

When I was a student following some courses on AI and expert systems, our professor used to insist on the idea that any statement concerning a corpus of knowledge could be just translated into one or more rules, and then an inference engine would take this body of rules (the “rule base”) and the combination would have the appropriate behaviour. What’s more he used to insist on the idea that statements came as an unordered collection.

It was appealing but never convinced me.

It doesnt always work in practice

That professor got a few contracts working with big players, and in one of them he was required to work with a real world problem in which it was practically mandatory to work with sorted data. So he tried to sort the data with some inference rules. It did not work well enough, it was too cumbersome.

As a way of cognitive modelling

Here are some examples where it’s you the human that is doing the inference:

1)You get an electronic device and you read that it has been designed to turn itself off if there is no user interaction with it for a period of more than 10 minutes. This is a battery-saving feature of course. It might take a little time to assimilate but it changes your behaviour with respect to that device, in that you stop worrying about how long it has been turned on. This is a rather deep ramification of the statement that it has this battery-saving feature. I really don’t see that as at all something that would fall out of a rule and an inference engine.

2)When you know that you can undo any action in a text editor, it frees up a good deal of your thinking and changes your behaviour.