Weak slot-and-filler structures are considered weak because they represent general knowledge that can be applied across various domains, rather than specific, highly constrained knowledge. This generality makes them flexible but less powerful for deep reasoning in specific contexts.