Assuming that the system supporting this career path works well, these will be the individuals with the most talent and interest in such an endeavor: young people well positioned to make the scientific breakthroughs that societies need to survive and thrive. But the current system squanders the creativity and energy of these exceptionally gifted young people through a funding process that forces them to avoid risk-taking and innovation.
The traditional peer review system on which scientists depend for federal grant support values biomedical research projects that are almost certain to "work," encouraging young scientists to pursue a narrow range of projects that closely follow the proven paths of their mentors. As a result, many scientists pursue identical research ideas, creating a competition to finish and publish that can value speed over quality. Worse, the innovation that is essential for keeping science exciting and productive is replaced by a great deal of "me-too" science: research that has little chance of producing the breakthroughs needed to improve human health.
Alberts adds one more dimension to the discussion: he criticizes the NIH's special innovator awards, designed to encourage less conventional projects, as being too little to have much effect on the culture of the biomedical sciences.
And he's right - why should bolder, higher-risk/higher payoff proposals be ghettoized to their own funding mechanism, one with a success rate that is 10 times lower than that of the R01 mechanism? As a new investigator, right now you have a choice: go with an unimaginative, conventional project that will, with a reasonably high probability, after a >1-year processes get you an R01, or put your grant-writing energy into an extremely creative idea that has only a 1% chance of getting funded. (Alberts writes that in a recent round, there were 2200 applications for 30 NIH Director's New Innovator Awards.)
Alberts suggests that we bump the number of these awards from 30-50 up to 500. I don't think that's enough. We need to overhaul the way bread-and-butter R01's get evaluated, so that near-certain feasibility is not the top criterion for funding.
Read the feed: