"Traditional", or "commutative", public-key cryptography has not come up with brand new ideas for a while. At the same time, there is an abundance of new ideas in the emerging "noncommutative", or "combinatorial", cryptography. However, "traditional" cryptographers are not eager to welcome these ideas and, moreover, in the last 2-3 years they opted to altogether boycott them. One of the reasons is that security models established in "traditional" cryptography are not applicable in "combinatorial" cryptography; there are other reasons as well. In my talk, I will discuss possible resolutions of this situation.