Model Theory - Applications to Artificial Intelligence?

[Bottlestein]

Having just passed the Qual. for Model Theory, I was wondering: Are the techniques applicable to fields outside of Math or Epistemology? Obviously, it must have some applications in Computer Science, but the people who use it are still in the Math Department here.

Does it have any application in Artificial Intelligence, or related fields? I know some people here work on it, so I was wondering if this is used in any way.

[Starglider]

Computer science relies heavily on numerous areas of discrete maths that are technically branches of model theory, but in practice you don't need to know (and compsci students don't learn) the more abstract underpinnings of set theory, types, formal logic etc to do computer science. This kind of maths is almost entirely irrelevant to artificial intelligence, excepting those people specifically working on automated proving tools and general maths support software (e.g. Mathmatica).

[Bottlestein]

^ Damn. I was hoping it could be applied outside. I guess in Math the applications are pretty interesting, but I like trying to have more options at this stage

At any rate, do you guys use more Psych - type stuff for AI programming, then?

[Starglider]

At any rate, do you guys use more Psych - type stuff for AI programming, then?

A significant fraction of AI work is closely or loosely based on human neurology. However basing higher level designs (non-neuromorphic) on theories of mind from psychology is currently out of fashion.

[Skgoa]

I am in a CS master's degree program right now. We use many different kinds of math all. the. time. - but only as much as we need to. I.e. we use those properties that we need at that moment and ignore those that we don't care about. It really depends entirely on how theoretical your work/research ist. I know many mathematics students who attend our more theoretical lectures, while some cs students took math as a minor.