There is an idealized view of academic research in which that research takes greater risks than industry, plans for the longer term, and is less concerned with the commercial success of a research effort than in the intellectual content of the research. On this view, academic research can take a longer view than industrial research and development, and can take on higher-risk questions since even negative results can add to the base of knowledge that is the goal of academia. When a research program does pan out, the results can be transferred to industry for further development, and the academic researcher can turn to the next big question. Along the way, graduate students are trained in methods of research and techniques of system design, and when they are done they can either join the industrial world or return to academia to continue long-term research and the training of the next generation of graduate students.
Those believe this will also clap for Tinkerbell.
--- Jim Waldo, "On System Design", OOPSLA'06