Monday, April 27, 2009

Herbert Kroemer & innovation

A 78-minute 2005 UCSB lecture by Herbert Kroemer, the German American 2000 Nobel prize winner in physics (for semiconductor heterostructures), has been recommended by his proud disciple ;-), namely Gene.

Aside from some cute details about the history of transistors, CDs, LCD, LEDs, and the evolving appraisal of the importance of related phenomena, he said lots of important general things that I fully subscribe to, for example
  • the key applications of a discovery create completely new things and possibilities, so it is wrong to judge new discoveries according to their ability to solve the old problems only and to supersede existing gadgets
  • in other words, more generally, in the most valuable research (and the most valuable parts of the learning process), the results (and applications) are not known to start with
  • the methodology allowing one to learn the necessary stuff and independently solve problems is more important than individual pieces of knowledge
  • the key physical ideas and methods are usually contained in a simplified model which should therefore be naturally understood first, instead of attempting to solve the most general problem at the very beginning
  • it is completely sane that the business world tries to pay for research influencing a finite future and the business world shouldn't be attacked for behaving rationally
  • however, the researchers must still be given the freedom to investigate questions that are important according to the pure scientific criteria of the research itself
  • an expert cannot care much about the fame of other people who offer their opinions on some research, its value, and the answers - he must be able to judge these opinions according to their content.

No comments:

Post a Comment