Path: utzoo!news-server.csri.toronto.edu!cs.utexas.edu!wuarchive!zaphod.mps.ohio-state.edu!unix.cis.pitt.edu!dsinc!netnews.upenn.edu!msuinfo!pleiades.cps.msu.edu!punch From: punch@pleiades.cps.msu.edu (Bill Punch) Newsgroups: comp.ai Subject: Critical Mass in a Lab (was AI genealogy) Keywords: genealogy Message-ID: <1991Mar15.155349.3064@msuinfo.cl.msu.edu> Date: 15 Mar 91 15:53:49 GMT References: <2571@enuxha.eas.asu.edu> <31363@mimsy.umd.edu> <5510@media-lab.MEDIA.MIT.EDU> Sender: news@msuinfo.cl.msu.edu Reply-To: punch@pleiades.cps.msu.edu (Bill Punch) Organization: Computer Science, Michigan State University, E. Lansing Lines: 49 In article , velasco@ngagi.ucsd.edu (Gabriel Velasco) writes: |> minsky@media-lab.MEDIA.MIT.EDU (Marvin Minsky) writes: |> |> >I noted in my well-known paper on Frames that Charniak's thesis |> >discussed the activation, operation, and dismissal of expectation and |> >default-knowledge demons -- and that "many of his ideas have been |> >absorbed into this essay." I've always had a lot of ideas of "my |> >own", but I've been really outstandingly good at learning from those |> >so-called "students". |> |> Dr. Minsky brings up a good point. It would be interesting to see how |> advisors are affected by their students. As he notes many of his |> students' ideas have been absorbed into his essays. This would imply |> that there is an even stronger advisor-student relation; strengthened |> by the influence that the students have on the advisor. This relation, |> of course, would not be evident if we only examine theses. |> Sometimes what appears to be obvious needs to be said explicitly, and I think Dr. Minsky does this when he mentions the influence of students on advisors. As a recent grad student and now advisor I think I cannot emphasize enough the influence that one's students have on opening new areas to explore, seeing things you missed etc. In fact, I think it is the "critical mass" concept that is so crucial for good things to start happening in a lab. When a group of "good people" (fill in your own definition, something like smart, inquisitive, hard-working etc.) get together with some broadly shared goal, they start to build on each others ideas and create a gestalt that is more than the individual pieces. I think the good labs build on this, and in this environment it becomes difficult to identify ideas with anything/anybody other than a time and a group. It is for this reason that it is difficult for me to trace "ideas" other than through a group of people together at some time. Advisors may act as the spark, as goal setters and even as focal points, but often times it is the critical mass of a good lab that pushes ideas forward. I have often stopped and tried to define that term "critical mass" more carefully, mostly so I could imitate its effects in my own lab, but is seems a difficult job. Do you need "smart" people, "hard-working" people, some combination of characteristics, compatible personalities, certain kinds of advisors, what is it that makes a lab "click" like that and start working more as a group and less as a collection of individuals? >>>bill<<< punch@pleiades.cps.msu.edu Call on God, but row away from the rocks. -Indian proverb