Subject: Re: ok, FSBs created schwag; now bet the farm on something better
Date: Wed, 19 Sep 2001 11:41:38 -0400

Adam Turoff wrote:
> On Tue, Sep 18, 2001 at 12:12:10PM +0200, Norbert Bollow wrote:
> > Innovation cannot be really be planned or managed,
> True enough, but then you continue:
> > but it tends to occur with the greatest frequency in environments that
>   ^^^^^^^^^^^^^^^^^^^^^
> > encourage open, participatory processes, but where at the same
> > time those who have promising ideas are given the freedom to
> > pursue their ideas.
> So there's a relationship between open environments and innovation.
> But it isn't necessarily a causal relationship.  Do open environments
> cause innovation, or is innovation likely to occur where commercial
> investment finds an open environment?

If I learned one thing from auditing some of my wife's biology courses,
it is that words like "cause" are a lot more complex than they look.
Here is what I believe a pretty full answer to your question is.

I believe that innovation tends to happen when people with sufficient
background are actively working on a problem.  It is only partially
predictable as to what goes into a sufficient background, or which
people are likely to make the innovation.  The more truly innovative
the innovation, the more likely it is that the right background for
the innovation involves unexpected components.  That is because many
of the best innovations come from finding unexpected connections, or
from finding unexpected ways to apply knowledge in a strange
direction.  (For instance what does painting have to do with building
better destroyers?  Quite a lot, it turns out, when the painter is a
naturalist who has been studying the principles of camoflauge...)

Innovation can happen in both open and closed environments.  We are
all more familiar with innovation in open environments.  However in
WW II there was substantial innovation in many fields produced by
people working in very closed environments.  Clearly innovation does
not depend on being in an open or closed environment.  Put the right
people together, give them sufficient resources, motivate them and
you are likely to get good results.

However in an open system you get more movement of people and
ideas.  Furthermore in a closed system there is more of a tendancy
for a boss somewhere to be able to give the order, "This is how we
are doing it." - which can freeze ongoing analysis of possibilities
before innovations are found.  And also once a new idea is found,
open environments encourage the spread of the idea, resulting in less
duplication of effort in reinventing the wheel.  Of course in modern
times this benefit is limited by the sheer amount of existing
knowledge people have to learn...

Also there is a scale issue.  Closed systems tend for organizational
reasons towards having fewer but larger research efforts.  That allows
you to tackle problems you otherwise could not.  However given that
innovation is generally unpredictable, and given that there are a lot
of very simple innovations we could discover fairly easily (both
assumptions are based on past experience), you would expect that
scattering research energy would be a good idea.  Indeed past studies
that I have seen strongly suggest that funding 200 research projects
to the tune of 5 million each gives far bigger rewards than funding a
single billion dollar project.

So while I don't think that open environments are a cause of
innovation, if you want to cause innovation it is a good idea to try
to create an open environment.  Unless, of course, the task you are
involved in (winning a war, building a better chip, landing a man on
the moon) makes it implausible to do so.

> Bell Labs and Xerox PARC are two examples of innovation occurring
> with commercial investment in an open environment outside of the ivory
> tower.  ENIAC and Project Whirlwind are two examples of innovation
> occuring within the halls of academia only because of the deep pockets
> of the US Department of War (er, Defense).
You can find innovation in all environments.  Motivated people with
sufficient background and open eyes practically cannot help but find
innovations.  (Many of which will leave more experienced people
aghast, but innovation goes hand in hand with mistakes.)

I once ran across a mention that the Romans may have discovered how
to make bullet-proof glass.  The story was that a glassmaker managed
to make a glass goblet which would not shatter when dropped, and
presented it to the emperor.  The emperor was pleased with this gift
and, wanting to be the only one with a goblet like that, killed the
glassmaker.  I do not know if the story is true, but it is neither
technically nor historically implausible.  And if true it shows that
even the possibility of death does not stop innovation!

> > A focus on Free Software research will create such an environment.
> I still fail to see a causal relationship here.  Maybe that's just me.
Not wearing a seatbelt doesn't kill you either.  Millions have done it
and lived.  It also doesn't cause you to live.  You can be wearing a
seatbelt and still die of a heart attack, if your car rolls over a
cliff and lands on the roof, etc.  (BTW the second is an accident my
brother had and he would have died if the boulder he landed on was 6
inches over.)  And finally most of us know people who have survived
serious accidents without a seatbelt.  (I would start with my mother,
who was thrown 20 feet while 8 months pregnant without suffering a
serious injury.)

However nobody who has looked at the question seriously doubts that it
is a good idea to wear a seatbelt, even though it does not cause life,
and the lack of wearing one is not the proximate cause of death.  Why?
Because it improves the odds.

In the same way I submit that having open environments does not cause
innovation, but it improves the odds of having the necessary
preconditions.  Therefore creating an open research environment is
good for getting innovation.  (*Taking advantage* of innovation is
probably a different story.)