The aphorism that innovation is the provenance of the young is longstanding, yet is increasingly less and less true. The mean age of inventors of “great inventions” and of the primary work of Nobel prize winners has increased about 8 years over the past century, with current mean age of highest productivity lying in the late 30s. Jones shows that this does not simply reflect longer lifespans, but rather seems to be related to increasingly long periods of training before research work begins. He presents a simple model of why this might be case, the intuition of which is simple: we are further from the knowledge frontier, the “low hanging fruit” has been picked, and production of new knowledge today requires both more training, and more specialization, to discover. This microdata comports with a well-known trend of declining productivity among R&D workers over the 20th century. The implication for growth is not good: since each knowledge worker needs to spend more time training, they will have lower lifetime contribution to the knowledge frontier (because they will have fewer productive years after training available), and hence growth of new knowledge will slow. Two trends may mitigate this effect: first, a Kuhn-style paradigm shift may open the door again to young researchers, as in the case of quantum physics at the turn of the 20th century, and second, new technologies, such as more advances computer assistants, may allow us to reach the knowledge frontier quicker than our forebears.