Comment by vharuck

>I guess I shouldn't discount the value of such a thing, once we get the "Sometimes it just lies" problem under control.

Considering the model doesn't "think" or understand abstract concepts, could we ever expect this?