The way biomedical research is organized has dramatically changed over the past half-century: are the changes for the better?

Biomedical research in today's universities is usually carried out by groups consisting of a leader and 5-20 or so trainees. This is in sharp contrast with past generations, when research was usually done by individuals or small partnerships of two or three who thought up their own ideas and carried them out themselves. Group leaders today spend their time in an office, on a wide variety of administrative tasks, and have little or no time left for work at the bench. I recommend that leaders try to change the system by forming smaller groups and insisting on reserving their own time at the bench. I suggest that trainees, before beginning their research, look for laboratories where groups are small and independent, with leaders engaged actively in research.