Why are some technological transitions particularly unequal and slow to play out? We develop a theory to study transitions after technological innovations driven by worker reallocation within a generation and changes in the skill distribution across generations. The economy’s transitional dynamics have a representation as a q-theory of skill investment. We exploit this in two ways. First, to show that technology-skill specificity and the cost of skill investment determine how unequal and slow transitions are by affecting the two adjustment margins in the theory. Second, to connect these determinants to measurable, short-horizon changes in labor market outcomes within and between generations. We then empirically analyze the adjustment to recent cognitive-biased innovations in developed economies. Strong responses of cognitive-intensive employment for young but not old generations suggest that cognitive-skill specificity is high and that the supply of cognitive skills is more elastic for younger generations. This evidence indicates that cognitive-biased transitions slowly unfold over many generations. As such, naively extrapolating from observed changes at short horizons leads to overly pessimistic views about their welfare and distributional implications.