John in fact is a mentor of mine. We first met, I think it must have been 17 or 18 years ago, when I'd gotten fascinated with the promise implied, the promise or threat implied, in the words "artificial intelligence." A mutual friend said, "Well if you call this number and knock three times, you may meet a bearded fellow in sandals." John was running a lab out in Palo Alto, adjacent to Xerox PARC, of which he was a part, and subsequently bought a pair of lace-up shoes, accepted a corporate title, and began commuting across the country. John is chief scientist of Xerox, he's head of Xerox PARC, he's a great spirit and a great inspiration. With that, John, live up to it.

John Seely Brown: Bill, thank you.

I'd like to focus on two issues, the first having to do with the changing role of corporate research, in particular how corporate technologists, intrapreneurs and entrepreneurs are beginning to create a kind of new matrix for innovation. And the second, involving a re-examination of the firm from the point of view of knowledge creation, which leads to us to the notion of a knowledge ecology - a concept that helps to situate the firm within a regional advantage while also suggesting the increasing importance of diversity.

So, let me start with the changing role of corporate research. Said most simply, we are expanding our role from just invention to invention plus disruptive innovation - disruptive in the sense of creating new markets (or industries) and innovation being invention actually implemented.

Why? In today's rapidly changing business landscape where existing businesses and business models are being constantly challenged, our mission might be most succinctly stated as enhancing the adaptability of the species, the species being, in my case, Xerox or whatever company is sponsoring the research. From this perspective, our research agenda must involve probing and scanning the periphery, pursuing discontinuous technologies, listening deeply to both our current customers and those who are not our customers, and challenging status quo beliefs within the organization. As such, we are attempting to be a kind of genetic variance for the species.

Key to transforming disruptive inventions into innovations is recognizing the need to look outside the firm for some of the complementary assets that are needed to bring the invention rapidly to market. Such partnerships or acquisitions not only help in improving time to market but, more importantly, improving time to volume.

The hardest thing I have found for us as researchers to do as we champion our inventions is to focus on a value proposition for the invention. All too often we have many value propositions, each half-baked, and to commit to one requires experiences that few researchers have had. Our solution in Xerox has been to create a Corporate Innovation Council comprising the key technologists and strategists of the corporation, and to use this council to provide seed funding for exploring and validating a value proposition for a potential market-creating invention. This exploration is done by coupling the technology team to business and market development folks whose sole job is to create new businesses. This team is usually funded for six months or so to create and validate a value proposition and then to prepare a business plan for either spinning this endeavor completely out of the company or for launching it as a stand-alone entity inside, or killing it. If it is launched inside the company, the employees who go with the internal startup are given 20 percent of the company and are also allowed to continue to draw on the assets of the labs - giving it a major advantage from purely stand-alone startups. After a two-year or so period, we determine if we want to spin the entity the rest of the way out (via an IPO or partnership) or fold it back into one of the current business divisions, or beef it up into an entirely new business division.

One of the unexpected byproducts of this process is that the researchers quickly learn that there is often as much creative problem-solving involved in transforming the invention into an innovation as there was in creating the invention in the first place. This social discovery has certainly transformed the desire of researchers to work with marketing folks and business strategists.

The second point also has to do with enhancing the adaptability of the corporation, but this time by increasing the learning rate of the corporation. How can a corporation learn from the naturally occurring events that are happening day in and day out across the corporation and how can it share those learnings (or knowledge) more rapidly and in a way that employees can internalize them and act on them? In other words, how do you create a knowledge milieu, perhaps enhanced by the use of the Web, that captures, refines and disseminates local discoveries, discoveries from the field?

For example, we have 21,000 tech reps - troubleshooters - out in the field repairing our systems. How might we capture their insights, refine and warrant them, thereby transforming them into actionable knowledge? Of course, an insight per se is not knowledge. The techniques we used to transform insights and opinions into actionable knowledge were mostly sociological and were closely akin to those comprising the sociology of science. In our case, the tech rep community used our Web-based technology called Eureka to create their own peer reviews of a tip or idea. Their peers would critique, refine and augment the idea, transforming it into a succinct, actionable piece of knowledge and then post that into a growing knowledge base for all of our tech reps worldwide. Crucially, the name of the originator (inventor) along with the peer reviewers would be attached to the piece, thereby building social capital for them and intellectual capital for the rest of us. This ever-growing knowledge base has turned out to be useful not only for the tech rep community but also for the sales community, the documentation community - since some of the discoveries of the tech reps reflected errors and outright misconceptions by the technical manual producers - and the manufacturing and engineering communities, although these latter two are now just emerging.

A small extension of this perspective suggests looking at the modern firm as a collection of communities of practice, where discoveries are continually happening within each community of practice and getting refined and swapped around through highly informal channels. Unfortunately, these learnings don't readily jump across the boundaries from one internal community to another. Explicit effort and technology that promotes boundary jumping is required - technologies aimed more at supporting the social fabric than the business processes of the firm. Intranet/webs look promising for such support and, if successful, could lead to not just organization knowledge sharing but to the firm itself becoming a miniature knowledge ecology.

We are finding the concept of a knowledge ecology, facilitated by the Web, to be increasingly useful, not only for rethinking the firm but for rethinking regional advantage. In a region it is common to think of its entities in terms of knowledge producers - schools, colleges, research universities, libraries, etc. - and knowledge consumers such as the firms comprising the region. Of course, the boundaries between producers and consumers are not always fixed or clear, since firms are increasingly producers of knowledge as well as consumers of it, and the same with universities. Also, within the firms themselves, one finds rich cross-pollination of ideas by the fluid movement of people within and between firms as well as by ubiquitous social interactions between people of different firms. Curiously, the Web - the purveyor of cyberspace - tends to amplify the advantages of the physical region by amplifying and sustaining the social interactions established physically and by amplifying the tendency to share ideas. This latter point stems from the fact that if one is of the Web culture, there is a much more fluid boundary between consumption and production of ideas. On and in the Web one tends to be both a consumer and a producer, sometimes intentionally, sometimes not. Consider just some of the new e-commerce sites that count on buyers writing reviews or providing feedback on the service they received. Although this shift is barely visible, I expect it will have a profound effect on how knowledge ecologies of the region will develop, but that is too long a story to go into here.

My last comment is, if you really look at the region, for example Silicon Valley, you have research centers, small companies, large companies, major universities, etc. It's very interesting how these all start to interplay, where ideas leak out, across and back in through these boundaries. To realize how this ecology gets nurtured by various forms of leakage, one must attend to the social context of knowledge sharing and informal benchmarking. Hardly a lunch or dinner goes by without my hearing (or overhearing) some new idea and wondering if it tops something we are already working on. That constant wondering keeps one both humble and attuned to the intellectual ferment of our knowledge age.