The current concern with a new digital transformation has once more placed the future of work at the center of the public and research agendas. As sociologists of work seek to grapple with its potential implications, it is perhaps salutary to consider the fate of their earlier attempts to forecast the future. The paper examines the empirical success of the major theoretical forecasts of the early postwar period and their successors in the 1990s. The last seven decades of research should make us careful about ready predictions of the direction of social change, attentive to the dangers of extrapolating from "cutting-edge" developments, and aware of the complexity of trends and the factors that can moderate them.