The paper provides a critical analysis of the idea of technological unemployment. The overview of the existing literature on the employment effects of technological change shows that on the micro-level there exists strong and positive relationship between innovations and employment growth in firms; on the sectoral level this correlation becomes ambiguous; on the macro-level the impact of new technologies seems to be positive or neutral. This implies that fears of explosive growth of technological unemployment in the foreseeable future are exaggerated. Our analysis further suggests that new technologies affect mostly the structure of employment rather than its level. Additionally we argue that automation and digitalisation would change mostly task sets within particular occupations rather than distribution of workers by occupations.