taking knowledge-based ai this semester, been reading ahead and seems interesting, basically a mashup of cognitive science and ai, how humans would approach problems. seems to give a bigger context into ai than my other ai classes have (ai for robotics, ai). i guess out of all the cs topics, i find ai the most interesting in general? in that it seems to patch together these relatively straightforward concepts (at least they start out seeming straightforward) and then magic happens and now you have a little thing that can figure out mazes or play games or what not. i wouldn't mind studying ai things more in the future but i suppose also i don't particularly feel creative in ai, or any cs topic really, and i suppose i've stopped feeling creative for awhile in my previously studied subject (music) although i kind of live in the opposite of a stimulating, creativity-inducing environment and don't really socialize,.. at all. starting to wonder what i'm going to do with this cs degree now that i'm starting the back half of it and thinking maybe evidence that ai is interesting to me might be helpful (which would seem to require some sort of creative motivation that i don't have atm). also, it seems that a lot of work in the field have applications that i would rather not get into.
one thing i find interesting about the class material so far is that when it talks about ai agent learning, it stresses small amounts of data (vs the large amounts of data a lot of ml methods frequently rely on). i kinda wonder how much "state-of-the-art" ai research is still focused on that "small data" assumption and also it makes me imagine like,.. ai agents with "we only use small amounts of ethically-sourced data" on the box... if computer software still came in boxes