Subject: Re: How to run an FSB R&D lab (maybe)
From: "Forrest J. Cavalier III" <mibsoft@mibsoftware.com>
Date: Thu, 21 Mar 2002 14:07:17 -0500 (EST)

> Even in an academic setting (MIT and CMU are examples from personal
> experience) there are often *also* contexts in which the mix is
> leavened by "outsiders" (usually alumni, sometimes staff) with Real
> World Experience(tm) to get the kind of thing Tom is talking about.
> 
> And note also that those environments (Athena, Andrew) are (1) worth
> looking at (2) take 10 years to become off-the-shelf, *because*
> they're heavily populated with *inexperienced* but creative and
> unfettered students, with "experience" pointing out more of the "well,
> that didn't work last time, at least try something unknown" instead of
> the kind of "here's a direction people want" that a business would
> expect to head in.

If universities are such a great model for creating saleable value,
then why have they been unable to hold tuition increases to the
rate of inflation?  Why do faculty go outside the university
structure to start up companies?

So, my take on this is:
  "universities are close to the model you want,
   except that average worker productivity is less than 0, and
   $145,000 lower than necessary in order to make it work."

Why?

If universities are a good model for high tech business, the model
must have revenue on the order of $120,000 per year per student.
They don't do that, of course.  Instead they charge the students
$25,000 a year to participate, even though they already claim all the
commercial and other rights to the significant work produced by students,
regardless of the proportion of investment.  (Aside: I was there when
CMU enacted their IP grab, and it definitely changed how much
software I wanted to create using their systems.  Anything that
I thought I wanted to keep and use later, I was careful to do
on my own computers.  So much for fertilizer.)

It is hard to even argue that they have education as their
mission.  Tuition is spent on maintenance, marketing, and
administration. The "educators" get a salary, but they actually
have to pay their own way through research or moonlighting (er,
start ups.)  Buildings are financed by alumni. (I better get
back on topic here....These complaints are not new.)

My point is that if you want to make the model work, you must change
it so that you can PAY the workers, instead of charging them.

In the case of institutions the size of CMU, the fruits that drop
from this flowering tree have to be on the order of 1600 hi-tech
workers (estimated enrollment of engineering undergrads) * $120,000
per year = 192 million dollars.   Of course you need that they are
all mature workers.  There are graduate students, faculty,
and staff necessary to the mix too, so budget accordingly.
You can't downsize too much, or you lose the fertilizing effect.

For CMU, I think Lycos is a good example of what Tom is
writing about.  I think it is an exception that doesn't come
along every year, and it generated less than $100 million to CMU
if I remember correctly.  (If it had sold at a different time,
it could have been more, and in the present hi-tech stock market it
would have been a lot less.)