In my talk I will try to explain the result of my joint paper arXiv:math/0605141 with D. Tamarkin and B. Tsygan in as popular way as possible. First, I will convince you that the notion of ``homotopy Gerstenhaber algebra'' is not so crazy. Second, I will recall algebraic structures on Hochschild cochains and formulate the main theorem. Finally, I will explain the idea of the proof referring to the well known bar/cobar constructions for associative algebras. Well, if the bar/cobar constructions are not well known then they will be well known after my talk.