The article covers the thirty-year history of machine translation (MT) in the former Soviet Union. The author stresses the necessity of developing MT in view of the changing social and economic conditions of present-day life. MT is considered to be an efficient way of dealing with the substantial imbalance between the volume of incoming information and the human capacity to process it. Two types of translation are identified from a functional-communicative point of view, namely literary and specialised translation. The author specifies that in his view MT cannot be used for the translation of literary texts but only of administrative, technical and scientific texts, the domain of specialised translation. The Linguistic Translation Theory forms a theoretical background for specialised translation.