Wired:
|
|
What’s
the single most important thing you think people need to understand
about innovation in research and development? |
JSB: |
|
Tools
drive science. Not theory, not experiment; it’s the tools.
And it’s this that has made the computer such an incredible
force for scientific innovation. For example, the ability of
the computer to crunch unbelievable amounts of information;
to design and fabricate micromachinery; to link disparate technologies
into networks; to create new materials with new properties;
and to visualize what’s going on in complex interaction
has completely changed the speed and nature of innovation. And
now the feedback loops are getting tighter and tighter as we
use computers to create new tools and then we turn around and
use these new tools to create yet more computing power and we
create more new artifacts and materials. It’s a virtuous
spiral. |
Wired: |
|
What’s
the human impact of these tools on how people think or on how
society functions? |
JSB: |
|
Obviously,
technology influences society and vice versa. Society and technology
form a co-evolutionary system. New tools have transformed by orders
of magnitude the questions we can ask and answer. I can now do
in one day what it once took me ten years to do, I can now think
about asking questions I would never have dared to think ask before.
Those questions were there before, but they were beyond exploration.
I’d never thought about asking them because they were impractical
to ask. Just as important, we see new communities of practice
growing around these tools that give us the opportunity to make
major strides by powerfully combining disciplines that were once
seemingly unrelated such as in chemistry and the theory of algorithms
as we see in the emerging field of nanotechnology. We have interdisciplinary
communities that otherwise wouldn’t have existed now asking
— and answering — questions they couldn’t have
meaningfully asked before. |
Wired: |
|
Just
how multidisciplinary are these new communities around new tools?
Xerox Parc has a pretty vibrant artists-in-residence program.
Is that a “nice-to-have” or a “need-to-have”?
|
JSB: |
|
Today’s
fast moving pioneering research needs artists and out-of-the-box
thinkers who are willing to challenge existing paradigms as much
as we need traditional scientists. Artists bring a different design
sensibility to tools than scientists and engineers do;the notion
of ‘composing’ in atoms and or bits is a deliberately
artsy metaphor. Many artists are also natural bricoleurs. They
instinctively use tools and materials that are at hand to build
new works, taking things out of an established and expected context
and putting it in an unexpected one. This way they make us think
in new and unexpected ways. That’s something science, too,
is starting to do—and scientists can learn a lot about working
this way from artists. |
Wired: |
|
How
does your understanding and philosophy of computation differ from
classical approaches to computation and engineering? |
JSB: |
|
We’re
trying to figure out how to work with nature and the world as
opposed to extracting the problem out of the world. You might
think of it as a form of judo where we ask how we can get the
world to do some of work for us. We start our inquiries by saying,
let’s not immediately go to the formal representation
because as soon as you formally represent it, you are taking
it out of its context. and sometimes the context contains the
key to a clever solution as Sherlock Holmes often discovered.
In
classical R & D, every time you come up against a road block,
you panic. The bigger the road block, the more depressed you
get, and the more you try to circumvent it. But when you’re
doing pioneering research, the bigger the roadblock, in some
sense the closer you may be to discovering something fundamental
about nature. So by going to the root of the barrier you let
nature drive you. If you actually get multiple disciplines together
working around the root of a problem, the problem itself pulls
you out of your own discipline causing you to fuse different
points of view that can lead to a fundamental reframing of the
problem. Good architects, for example, constantly figure out
how to transform constraints into resources as do edge designers
and pioneering re searchers. Consider, for example, how some
of the most recent research in creating an AIDS vaccine is,
I think, trying to find ways to leverage the rapid mutation
of the virus as a resource. |
Wired: |
|
That
goes right back to the issue of how artists use tools versus scientists
and technologists. |
JSB: |
|
Artists
use tools in a playful way, because they don’t necessarily
have to solve new problems. Conversely, the biologist or in my
case material scientists begin to understand we can compose matter
by bringing bits and atoms together in novel ways. For example,
the beautiful work by Mark Yim and his group on building polybots—modular
robots that can reconfigure themselves for the task at hand suggests
what might be possible as their modules shrink in size using MEMS
(micro electrical mechanical systems) therein creating something
they call digital clay. |
Wired: |
|
How
does that help? |
JSB: |
|
|
It
started to bug me this morning while I was thinking, ‘Why
don’t we have more avante guard scientists?’ We have
an active ‘artists in residence, where the artists are almost
always avante guard artists — artists who are skilled at
pushing the extant genres of their trade. But if you are pushing
or transgressing the boundaries what guideposts do you use? The
answer is simple—taste! A personal but well honed sense
of aesthetics. |
Wired: |
|
Do
you really think of yourself as a composer, or as an editor? Because
a composer has to start with a blank sheet of paper while the
editor, at least, gets to start with something. |
JSB: |
|
The
composers of my world bring atoms and bits together, which is
[where the next big action really is]. When you bring atoms and
bits together, you are really composing smart stuff (i.e., materials).
I think creating new forms of matter and media is a different
game, driven by a mixture of composing and blurring—blurring
the boundaries of computation and matter. |
Wired: |
|
So
“composed” is a more artful word. |
JSB: |
|
Yes,
absolutely. For example, I think the ability to assemble and control
huge numbers (even zillions) of micro machines each driven by
local rules with an ability to wirelessly communicate to its nearest
neighbors will create a whole new genre of self diagnosing, self
repairing systems. For a trivial example, consider some new kind
of smart stuff for building walls where the smart stuff can cancel
sound by reversing the phase of the sound wave. So the wall becomes
perfect at absorbing sound by canceling it . Although such material
does not exist yet, similar ideas are being used to cancel buckling
moments of steel columns. I could also exemplify this through
some of our radical new approaches to printing but I will refrain—we
have learned some lessons from the past. |
Wired: |
|
Why
did you pick “compose” rather than “author”
or “sculpt”? |
JSB: |
|
Sculpt,
I didn’t think of it, but we certainly do a lot of sculpturing
with some of the new deep etching techniques. But compose just
seems to me to capture the spirit of bringing atoms and bits
together. It is just the sense I get as I walk the halls of
PARC And composing reminds me that (like some composers), we
are conductors, too, orchestrating different practices and pieces
of material so that they work together. |
Wired: |
|
What
gets you more excited, knowing you can find a charismatic young
research leader or a charismatic new tool? |
JSB: |
|
In
today’s world,I would go for charismatic tools. |
Wired: |
|
Is
that a politically correct answer or do you really believe it? |
JSB: |
|
No
— I believe it. It’s been my experience although,
of course, there is a deep interplay between the two. After
all, who created the ‘charismatic’ tool in the first
place? Isn’t this question a bit like the chicken and
the egg kind of question. The real key is how to create a charismatic
milieu that combines leading edge tools with a culture of being
an edge designer. |
How Do You Tell the Tools Apart?
|
Wired: |
|
What
makes one tool better than another? |
JSB: |
|
Obviously,
different tools serve different purposes. I like to think about
the Q factor of a tool. The hammer can be used for all kinds of
purpose many of which it does reasonably well. On the other hand,
the electron microscope has a limited set of uses but it does
those extremely well. It has a high Q factor. Use it for how it
is meant to be used and it is incredibly good; use it in ways
that it was not designed to be used and you have problems. |
Wired: |
|
Ahhhh,
Q-factor. Isn’t that a term used in electronic circuit design. |
JSB: |
|
Yes,
it’s a term often applied to filters, etc. I use it here
because it has a nice metaphorical meaning especially for computational
tools. Consider the original email systems, by just transmitting
unformatted text everyone could use it since the protocol was
so simple — almost nothing had to be pre-agreed. The same
with the original HTTP but as we move to new generations of email
and http, we are increasing their Q factor. The new protocols
help us do some things incredibly well but can complicate the
interpreters we build and may work against novel uses of these
tools along dimensions not honored by the protocol extensions.
|
Wired: |
|
We
have this world where CORPORATE research is A MAJOR source of
innovation. Is there any way that we can transfer this on a larger
scale to many smaller individuals? |
JSB: |
|
For
some kinds of activities, yes! Consider various OPEN SOURCE movements
such as Linux, IETF or even the original MOSIS system for bootstrapping
the community mind about VLSI design. |
Wired: |
|
Do
you see the open source philosophy being an important ingredient
of innovation? |
JSB: |
|
Absolutely!
I consider, for example, Linux an amazing achievement in how
to tap and support the creative juices of the community mind.
We need to better understand what makes it work, in what other
circumstances will such an approach work. For example, would
Linux have achieved it current successes without a master architect
carefully adjudicating what goes into the core and what didn’t.
It’s easy to think of Open Source as unstructured or unorganized,
but we have to realize that intriguingly being able to say no
is perhaps the most critical skill in designing complex systems!
How might we lift the learning's of this open source community
to one that might, for example, create educational material.
How do we better understand the role of social capital in fostering
such movements? How is social capital traded off against financial
capital and need it be? |
Wired: |
|
What
do you find provocative about “open source” as a
tool or a medium? |
JSB: |
|
As
we discuss in our recent book, The Social Life of Information,
we—as technologists—have tended to focus on information
and individuals leaving out context, communities. Open source
is a beautiful example of how the expert system often lies in
the community mind. With the rise of the internet and tools for
supporting virtual communities we may now be in position to really
leverage the community mind. Open source may also give us a way
to crack the robustness problems of really complex systems. In
Linux, for example, you write code to be read by others as well
as executed by the computer. Writing code to be read is a great
form of community hygiene. And when code is meant to be read by
others, it has its own social life, so to speak, and as such it
gets picked up by the community and used in all kinds of new ways.
Pretty soon we all become bricoleurs and then the community mind
becomes a new kind of platform for innovation. |
Wired: |
|
How
much do we have to know about making judgments about consciousness
to do research? |
JSB: |
|
I
said earlier that taste, a personal sense of aesthetics are part
of what artists can help scientists learn. Because taste and judgment
are critical parts, but easily overlooked parts, of scientific
practice. Their part of the tacit understanding that scientists
bring to their work. We tend to think of scientists as automata—brilliantly
logical, but purely logical, working on explicit knowledge. But
great science and research involve intuition, from choosing the
right problem to address and judging the right way to address
it, to recognizing what counts as a worthwhile answer. |
The Best Tool is the World. Nature as Tool.
|
Wired: |
|
Yes, as I said at the beginning of this conversation I try to
use the conceptual lenses of judo. That is, how do you let the
world do more of the work for you by interpreting the invisible
physical and social forces at work and leveraging them rather
than ignoring or fighting them ? When I was a purebred computer
scientist, I was taught to take the problem out of the context
and represent it, compute on it, and put it back in to the context.
So knowledge representation, [data structures], and algorithms
were the sole coin of the day. Now what we’re beginning
to see is, if you can actually build systems that interact with
the world, you can let the world do more of the work. When you
come to composing the new types of structural smart matter, you
actually have now a chance to figure out how you work with [physics]
— not against it. You let physics of the situation do some
of the computing? And, of course, our discussion of various open
source movements is another example of leveraging the social world—or
actually enabling the social world to leverage itself once you
see how the invisible social world of information works. |
Wired: |
|
Above
you gave some great examples of the social but can you give me
a concrete example of leveraging physics to help solve your problem. |
JSB: |
|
Sure.
In transforming how one might do high speed color laser printing,
we wanted to create a new kind of optical system for spraying
light on the photoreceptor belt. The idea was to create a tiny
chip that had thousands of lasers on it and on top of each laser
to build a microscopic lens that could shuttle back and forth
a few microns, steering the laser beam as it moved. But how could
we make these lenses, each less than the diameter of a human hair.
Grind them or grow them. The technique our guys finally converged
on was so simple it was shocking. In essence, take some of the
lensing material, confine it to a tiny circle and heat it up.
WhamO. The surface tension of the liquid causes a lens to form.
Wonderfully simple. Let the world do the work. Elegant, yes and
extremely practical. As Paul and I say in our book, the way ahead
is often to look around. |
Wired: |
|
One
last question. Do you see any fundamental shifts on the horizon
for R&D Centers? |
JSB: |
|
Absolutely.
I think a fundamental shift is happening stemming from finding
ways to leverage knowledge ecologies such as Silicon Valley.
These ecologies are great at exploring many ideas in parallel;
they are so great at doing this that you find that there is
a shift underway that turns much of the classical R&D into
A&D — that is, acquisition and development. Indeed,
that is one of the reasons we have been focusing more of our
effort on pioneering research that explores the whitespace between
disciplines and between the arts and the sciences. For such
explorations you need to have a critical mass and a rich ecology
of disciplines and tastes all working shoulder to shoulder.
That is difficult to do justify in the valley and it is a unique
role for a certain kind of corporate research. |