Interview with Doug Engelbart, 1999

In 1999, I wrote a new afterword to the MIT Press edition of Tools for Thought, originally published by Simon & Schuster in 1985. I re-interviewed Doug Engelbart, Alan Kay, Bob Taylor. Here is the unedited text of my interview with Engelbart (which took place at his home in Atherton).

Interview with Doug Engelbart

By Howard Rheingold

July 28, 1999

H:  Exactly.  I am amazed at your perseverance in continuing to further the project you started so long ago.  So just in general, in terms of agenda-setting here, I wanted to briefly get the history of what you’ve done since then, and then talk about how it looked, how the future looked back then, what your intentions were, and how things have turned out, and how the future looks now, and what your intentions are.  And I guess, I don’t think we talked about CODIAK back in 1983, so maybe just a brief description of how you’ve come to kind of reformulate what the original program was.

D:  Well, it’s sort of reformulating or reclarifying it, it hasn’t changed very much.  Just thinking more about it and developing newer concepts about how to think about it.  Let’s see, in ’84 Tymshare was bought by Mcdonell Douglas, remember that?

H:  Yeah.

D:  Yeah.  So then we all became Mcdonell Douglas employees, and Mcdonell Douglas was trying to build up an information service company, ISG or something they called it, Information Services Group.  So that they’d have some experience before that with their CAD system, making it available to other people, so they wanted to go into the whole business.  And it failed after 5 or 6 years.  And principally, I think, because they were used to doing business in mostly government-oriented defense planes, etc.  It’s very different environment when you start getting out there, very different product lines and all.  But it was very good experience for us because this enabled me to start going and interacting with the people inside of Mcdonell Douglas doing design, manufacturing, and support for aircraft.  

And up to that point, Tymshare had been quite clear about my not talking to the customers.  I want to be careful about what gets out in print about some of that, because it’s very easy to paint people as being short-sighted and reactionary, etc.  But if you really look at the prevailing paradigms at the time, you realize they were in line with the paradigms and I wasn’t.  So you just can’t pick out individuals and blame them.  Sometimes I think somebody ought to do a lot about the evolution of those paradigms of the time.  There’s a whole thesis about, if you don’t find out ways to evolve our paradigms faster, then they get farther and farther behind the actual future that’s emerging.

H:  Yeah, that seems to be the situation we’re in, in fact.

D:  Right.  So that became so clear to me that it became part of the strategic concepts about, well, what can you do to contribute?  The underlying thing for me has been, how can we get collectively better at handling complex urgent problems?  Because they are getting more complex, more urgent, more rapidly, etc., and our rate of improving our ability to do that complex stuff has got to keep up.  So then it became clear to me that evolving paradigms is very much a part of that whole pattern.

In parallel with that, the commercial aspect that got started at Tymshare was doing really very interesting things, helping mostly military groups, because they were the only ones who had access to the Arpanet, and the verge of the Internet.  And then inside Mcdonell Douglas, that’s where I was getting most of my experience.  So, it just really then came home of things, how the standards for how knowledge containers have to be standards so that the engineers can talk with the manufacturing guys, can talk with the suppliers who also have engineers and salesmen and planners, and with the contracts people who have to negotiate and support contracts, and with the manufacturing people that, just you realize it’s a bigger and bigger set of different types of people throughout the world that are going to have to work together, until you just realize that something about an open system of knowledge containers just has to be there, or you’re not going to get the kind of complex support that you want.

So that’s where the idea of an open, hyper-document system emerged.  And we several times almost got some real muscle behind the whole movement.  Once almost within Mcdonell Douglas, that had what would now be the CIO, said, “Well, this all sounds kind of interesting, but you guys go out there and make a good example.”  So he picked a project.  “If you can go get those guys interested and involved, then you can come back to talk to me.”  So it took a year, and this was a project having to do with supporting the Exploratorium, planning a design for aircraft that would fly 100 miles high. 

And so Mcdonell Douglas had part of that, they were the chief project leader on part of that, and so we could actually go start trying to get people to use the FOGNET??? system.  And we soon got told, “Well, we’re under a lot of pressure so it’s all right with us if come in and try to get people who want to use it, but the thing is, you can’t take any of their time for training, they’re all busy 6 days a week, 12 hours, something like that, and you just can’t.”  Our jaws dropped, but the guy was totally serious.  So here’s a Vice President saying we have to go get those guys interested and the head of the project saying we can’t take their time for training, and so somehow or another we just sneaked in.  People got to using it and they got very turned on and they found ways do project reporting and management that were just very novel and excited by it, and so we went back to the corporate vice president and said, “Okay, they’re turned on, now what?”  And he says, “Well, I’ve been talking to other people and I’ve talked to IBM and to DEC and to Hewlett Packard, and none of them knows anything about all this link shit you’re talking about”, he says, “So I’m not going to involved in that, we can’t do something ourselves that our suppliers don’t do, so forget it.”  It was so disappointing to one of the guys in there, he just couldn’t accept it.  He kept saying, “But this could make a terribly big difference in the whole effectiveness with which we can develop and manufacture aircraft.  It’s real.”  And the guy says, “No, listen, I’ve made up my mind, so shut up.”  And he kept talking, and he finally stood up and pointed at the guy, and says, “You better find a place in this company to go hide, because I’m going just squash you if I see you again or hear about you.”

H:  Wow.

D:  That’s big corporation stuff.

H:  Yeah.

D:  So the guy’s probably one or two levels below in the management chain where he would have been **** tending to be, just because of that.   And once before a guy had to resign from the Air Force, who was doing that, telling a colonel, “This is something you’ve got to listen to.”  Finally the colonel just said,  “Boy, you just ruined your rating, your evaluation rating.”  

H:  So, this is like 1984, 1985, 1986? 

D:  That was probably ’86, ‘85-’86.  Then there was another really exciting thing in the commercial side of the ex-Tymshare thing.  We’d been approached by people in the Air Force communication command who had charge of all the data processing and inter-communications of all the U.S. airbases around the world.   And they were trying to reformulate how they should give the computer support ??? out.  And so they [???] was part of it, and they made up a concept of operations, and started looking around, and they found us, and they said, “Gee, you guys have got just what we’re looking for.  You want to get up some prototypes?”  So we ended up with four mainframes sitting on their center in the US, actually in-use prototypes of this thing for more and more of the actual way in which they worked around the world. There was some special things added to that.  For instance, if a query came from Congress that hit them and always they had to pass it around to different people to see who could generate the answers, boy, now they would e-mail a journal and a separate system where you could find out, some guy would have to respond to it and pass it on to the others, so that anybody could stop and find out what the status was.  “Oh, it’s hung up now with this guy,” or, “he’s reported that,” or they could just track that on-line.  It was so much superior to anything that they’d ever seen.  They were very enthusiastic.  

So they generated a RFP for re-equipping hardware and software at all the bases.  We didn’t stimulate that, that’s what their plan was anyway.  So they built into the specs the kinds of things we were being able to do for them.  

H:  So what happened?

D:  We tried to go get DEC to bid with us on that, but they had their own office automation system, called “All-in-One”, or something like that, and those guys just said, “Wait a minute, wait a minute, you can’t do that.”   So, it flip-flopped around awhile, and then DEC was going to not cooperate with us, and so AT&T won it for $93 million for a system that they never could deliver the parts as far as [that].  DEC lost.   If either one of those had taken it and done it, it would have been bizarre.  Anyway, so then in the late ‘80’s, it just became so clear that we’d have to get cooperation from a number of organizations to get going in this because it was much bigger than any one system could go, and we’d never…  I approached IBM, and DEC, and HP, and said, “What about a product line of this sort?”  And none of them would buy, they weren’t, they were sort of empathetic with us, they’d talk, but this just totally wasn’t the market picture in that world.  So that’s when we formulated the Bootstrap Institute idea.  We set up an independent flag, and say, “Here’s a strategic approach that collectively organizations could get together and go after it.”  And that’s what we did, ten years ago.  So that’s had a number of chances that we’d take off and do something and how we couldn’t get them to work, and meanwhile these concepts like CODIAK emerged, as sort of a concept, by having a dynamic knowledge repository come out of it.

H:  Is it time to explain CODIAK?  

D:  Sure.  The acronym derives from the expression “COncurrently Developing, Integrating, and Applying Knowledge”.  Each of those terms is important, and it’s just one way to pile together the set of capabilities the future organizations are going to be a lot better at.  We also started using the term, “collective IQ”, as just a general way to say, “this aspect of organizations that is there to go out and improve”, and so CODIAK was trying to isolate some of the top level capabilities that you’d need to do that.  The developing and integrating and applying are all parts of it and any one of those terms embodies a lot of stuff that’s out there today, but as far as I know there isn’t any kind of coherent collection among those, especially the integration term.  You can have all kinds of dialogue, and all kinds of computer-aided design and planning and everything else in there, but the integration aspect is sort of…  We used to liken it to sort of this electronic handbook, where you assume that there’s any given??? body of dynamic pursuit from an organization going after it.  How do you go find out what’s dynamically up to date and integrated, so that the vocabulary is coherent and the open issues, what their statuses are, and decisions and plans that are established.  If you want to go back and question any roads you can trace back and find out the set of issues that arose and how they were resolved, and plans and the ideas.  So all that stuff is recording the dialogue that went on and the intelligence collection of what’s happening in the outside world, and integrating all that constantly into that handbook.  And then we had a MIT friend who came up with “organizational learning”.

H:  Peter Senge or Tom Malone

D:  Any of those, stuff came out like that.  To us they were talking about prior work to CODIAK, and organizational learning is, if you want to develop it and integrate it and apply it, that’s where you’ve got to learn it, in organizations.  So we hadn’t isolated those bits of things in there, but they were to be included.  The concurrent part of that is another surprisingly important, because you could also say collective or collaborative development, etc., he says, “Oh, sure,” but then he says, “Wait a minute, if we got a big project inside of Boeing, we’ve got to have our own CODIAK process going, our own knowledge repository ???  But anyway, within Boeing as a whole, they’ve also got other things going that they’ve have to work on too.  A whole picture of Boeing has all these concurrently developing knowledge activities, activities going to focus, each of them has to have a coherent CODIAK process going.  But you also have to have it all together going.  So concurrently all these things have to be concurrently coherent, as well as independently coherent.  That makes a real, real challenge on this interoperability and the conventions and the vocabulary, etc., that are employed doing this CODIAK work, the concurrency thing.

H:  Well, you know I remember once when you gave one of your slide shows, you talked about the way organizations do things, and the way organizations improve the way they do things, and the way organizations improve the way they improve they way they do things.

D:  That’s the big thing that evolved in 8-9 years ago.  

H:  And of course now there’s a lot, knowledge management is a big buzz word in organizations that do have people whose job it is to look at the way they improve the way they do things.  A lot of what you were talking about at the beginning is kind of the less visible, it isn’t hardware, it isn’t software, it’s the way people use it, the HLAMT stuff.  And after all these years, people are starting to really integrate that into their business, the non-visible part of technology.

D:  We start being awfully explicit too about the difference between the tool system and the human system.  The human system just embodied all the practices and conventions and vocabulary and skills and roles, etc.  We did talk about the co-evolution of those, that’s the critical thing, and that in the end that’s always happened, after technological innovations occur.  But then there’s a subsequent co-evolution of the whole environment you’re working on has changed.  Very seldom do you come in and just automate something that has no other effect.  So the big thing about augmenting rather than automating is being able to get the co-evolution.  We talked about this co-evolution frontier and that’s become a big part also.

H:  Well, I remember a couple of years ago at meetings at Sun one of the things that can happen historically, since we’re still talking about the history here, is that a lot of the people who had been your students or been involved in projects you had done over the years were now chief technology officer at Netscape, or at Sun, people like Jeff Ruhlson???, who really understood what you were talking about because they were there at the beginning, but here, what 20 years had passed?  25 years had passed?

D:  Let’s just say, it doesn’t necessarily follow that they were there at the beginning and that they understood it.  Anyway, it was very hard to communicate, it really took much understanding.

H:  Yet here we are years later, when you’re trying to get the Bootstrap Institute going, with people at a fairly high level at these companies who are very interested in doing that.

D:  That’s very, very true.  I’m just commenting on, back there in those days, there was almost zero dialogue about the underlying concepts, that the technologies were fun to do and exciting and foreign.  There was a lot of dissension about what we should be doing next, and there’s a lot of stress in the [???]  The emergence of mini-computers, and people all excited, “Let’s move on to mini-computers,” etc.  I was telling them, you’ve got such limited capacity there that you couldn’t possibly be exploring the kinds of things we’re exploring here.  And so there was a branch and the road and it took so much to explore how you can really do this, time-sharing and intercommunicating, the collaboration would be there

H:  I remember when we first talked, the Macintosh was, the world was beginning to see the GUI, and you were pretty adamant about the fact that these machines were not connected to each other, so you lose the important part of it.  Individuals could augment their capabilities, but groups could not, and had all these decades of experience with Internet by then.  What’s happened since then is that the Internet has become so ubiquitous and PCs have become so connected that kind of in a backwards manner that couldn’t have been predicted, we do now have hundreds of millions of machines that have that graphic interface that you pioneered and are connected to each other.

D:  We want to come back some time to the GUI, because I don’t like it.

H:  Okay, tell me why.

D:  (laughs) I find people sort of connect me with that, but it’s like actually when we first started building our systems, then we came to the point of talking about, “Do we use menus or what?”  By then I had enough experience with that keyset to just take a look and say, “In the time it takes me to move my cursor up to click on the menu item, I could hit two or three characters with my left hand on the keyset.”  And a lot of information is involved in making that selection, it’s much less than that much information.  So that’s not efficient.  In fact, if you are doing parallel at pointing at things and saying what you want to do with your left hand, you’re much better off saying what you want to do with your left hand than what you want to do it to with your right hand, and not try to have just that one point to vote???  It’s much more efficient.  So we just bypassed all together the kind of command feedback dynamically.  The computer would sort of guess what you’re saying and put it up there so fast that you had to do minimal input.  Anyway, it just let you start developing a big repertoire of verbs and nouns, more nouns, because you could explicitly express it.  And so it started to be a command language, and when you look the GUI as it is today, and I just think it’s like a very primitive pidgin English.  It’s like grunts and points.  And so I just, it’s bothered me greatly all these years.  

H:  Do you think that evolutionarily it isn’t going to happen?  Or do you think that 5 years or 10 years or 15 years from now you’re going to see something like the combination of the GUI in a command line that you’re talking about?

D:  One of the things that we did as a design principle starting in the late ‘60’s was say that we want to provide for a range of user interfaces, depending upon the kind of skill that a user has.  So we had the profile of verbs and nouns that you could use, could be something set for you.  So if you’re a beginner, a very simple subset, and you could just do a lot of things.  Then as you wanted to learn more and more, you could have them added and they wouldn’t have to be popping up, surprising you with things you didn’t understand.  And so this is in our picture of what an open hyper-document system would be, very much the idea of a range of user interfaces.  If people wanted to use today’s GUI, sure, fine.  If they want to start migrating to have a more powerful vocabulary and set of processes, etc., they’re free to do that.  So that just to me is about the only way you’ll get real evolution out there.  It’s not as if you want to give somebody a really high performance capability, they’d have to do a really simple isolated thing. 

H:  Well, the other thing that happened was the Internet and the Web, which does have links, but not exactly the way you had described it.  So what’s wrong with that, in the way that you…

D:  It’s a start, which is very, very important, that [we’ve] found some sort of combination that is simple enough, usable enough that it’s caught on.  So that’s a terrifically important contribution that happened.  But what is also very, very important is that people don’t get stopped by that.  And how do you get evolution from there that’s going to be effective?  Not having evolution such that you’re either using this whole system here, or you’ll have to transfer over to something very different.  So this is why the specifications for what we think you have to start giving in the open hyper-document system prototype that starts evolving is that is something that the architecture facilitates and use facilitates evolution.  And if high-performance people in the organization have to use a totally different system from the others, this won’t work either.  So he says, “Okay, you get it so you can have a range and so you can train high-performance teams of people who can just do a lot of effective stuff in there but they’re working on the same knowledge base that everybody else is.”  But that means there’d be properties embedded in the documents that the…I started using the term “pedestrian user”, that just ought to learn how to do this thing unless they’re interested in learning more.  Okay, user, GUI, whatever, but they won’t even know about these properties that may be embedded in the document or embeddable in there, they won’t have to.  So there are things that the high-performance ones can start depending upon.

H:  What were the most important general aspects of the open hyper-document system that we don’t have now with the Web that we need to see?  For example, you can link to a page, but you can’t link to a word or sentence or a paragraph, so that the granularity…

D:  I published in ’92 papers talking about that, that listed a bunch of those things, 

H:  What are the three or four most important ones, do you think?

D:  Well, integrating the editing of the browser.  It seems ludicrous that you do your editing over here, then you come over here to do your browsing.  That while you’re really trying to develop and compose some complex document, you’re just going to be moving all around it.  And want to view it in all sorts of flexible ways, and edit it, and whatever you do at any time no matter what view you have should be a relevant, ???  document.  So that’s one.

And then another one is that every object needs to be addressable intrinsically, not that you don’t have to go put a special tag on a paragraph or something, so it has a pound sign addressability in the document.  Another one is the option of viewing, the different ways you can view a document, knowing that it’s optional views, simple ones like the first line of every paragraph, another one would be if your optional can arrange hierarchically in your document, but you only see [???]  In HTML you can have H1 and H2, headers 1, headers 2 or 3, but they have no generally accepted bearing on hierarchical level, they’re more just for the addressing of how you wanted the font.  So … those conventions then…  if you wanted to start doing that.  So how do you … anyway, those are a few.

H:  Okay.

D:  Then the links need to extend the syntax of the link, so it’s not just an address, or even an address that includes pointing to a specific object, but also it should have syntactical place which you can specify the view when the reader follows that link and gets there it shows that particular view.  Then the things that we didn’t have yet but were waiting for were the kind of things where the contention in my troops would be that I’d like to do this and they could see that it would be valuable and it would be much more important than redoing our compiler or something with our money.  But the thing about back-link management of things about, whenever things are pointing to … you’d like to know about them, and there are various kinds of things you’d like to know, and this would lead into, say, “Oh, also in the syntax I would like to put ‘type’ in”, so we’d have conventions about what type of link I’m making when I point to that, and things to learn about that so that they help in analyzing it.  Is it a supportive thing?  Is it another kind of critique?  Does it give you an optional idea? Is it just using this as an example? Are you familiar with the IBIS and GIBIS sorts of evolution?

H:  Is that library systems?

D:  No, IBIS was Information-Based Issue System.  GIBIS was Graphical Information-Based Issue System.  Who’s the guy that wrote the summary thing about hypertext systems back in the 80’s and then went to Austin to work with a consortium?

H:  I vaguely remember that.

D:  Anyway, he developed a graphic one for that and then left to form a company.  It was a very neat way in which you could, in the dialogue about something that’s going on, you could keep track of issues and they could be explicitly tracked, to see when they were resolved and how one issue diverged or split into two.  They just had graphical maps of the issues they were tracking.  Just very neat.  But Stangwall???? by itself had a hard time to get going, but that’s very definitely the kind of thing that would work in this typing, because we did typing there.  And also Flores and Winograd’s system for, I can’t remember the terms that they used, also depended upon typing.  So the fact that having syntax in your link that had both ViewSpec and typing would be very important things to learn.

H:  ViewSpec and typing.

D:  Yes.  Somebody else who’s been prominent in hypertext did his thesis, and it didn’t work and he started talking about typing a long time ago too.  So all those things are part of what I call back-link management.  

H:  All right.  So this is a little bit about what I meant about how did things actually turn out as opposed to how you foresaw them, something that I’m talking about with everyone I interviewed??  back  then.  Other than the problems with the GUI and what needs to happen with the Web, I guess the other part is integrating knowledge management part of the system into the organization.  We’re beginning to see that with companies like Andersen and others who now have chief knowledge officers whose job it is to try to do part of what you were talking about way back when.  

D:  So there a just a couple of other things in the basics, but instead of trying to exhaust it there, we say, “well, since we started the Institute we’re getting more clear about the improvement process, about saying, ‘yes, there are activities that are there to improve your everyday work, and let’s clarify them.’”  That’s what we call the “B” activities.  And then what you have to realize that in times of rapid change about how work can get done, you’ve got to consider how you improve that improvement.  So this was another “C” category of activity for any organization.  Any activity was how they did their work today, all the skills, the capabilities, the vocabulary, the whole way they did it, but this has got to be improved, and so there have always been activities that changed, and improved.

So it’s just a matter of labeling them when you talk about this “D” activity.  They’ve always been there, and people are realizing they’ve got to spend more and more in there because things are changing.  But to move also to say, “We have to improve our capability to improve.”  What do people have to know in order now to really come in and build the Web into the environment, whatever project groups you have or there have to be some out-source consultants who’ve learned how to do that, so there has to be the activity of improving the improvement process.  Which turns out to be a very important concept in strategic thinking about all this.  It’s important in extending.  You can go to an organization that feels very progressive and saying, “How much are you spending on ‘B’ and ‘C’?”  I haven’t found one yet that has an accounting system that they can extract those numbers.  They say, “Well!”  You’ll find if you ask them how much they’re spending on product development, “Oh, boy, in our industrial niche the average is 8% and we’re spending 10%, [???]”

So these are the things that have gotten important enough that people have started isolating them so they can talk about it.  But just improving your capability to work hasn’t been clarified like that.  But the “B” and the “C” are very important strategically.  And then you say, “Well, what you do in “A” is sort of proprietary and we’d just as soon not reveal all that.”  What you do in “B” sometimes explicits, “I’m spending this money to improve this aspect of my manufacturing”or this aspect of something that you might want to keep quiet.  But what you do in “C” isn’t, it’s much more general.  And in fact, you’d rather not have your own independent picture of “C”, you’d like to be working with the model of how the world’s moving and where it’s going, what are the prevailing tools, methods, techniques, capabilities, best practices, and that’s the kind of thing that that often you out-source to consultants.   But the question is, how do you generate a picture when the world is changing so fast?  At this frontier of all the kinds of technologies, and all the kinds of new practices that these technologies can provide, makes a future that’s the frontier whose boundaries are much farther away.  And it’s a much, much, more, I guess you could say, turbulent frontier than anything we’ve ever faced before.

H:  So the Bootstrap Institute started a consortium of people in different companies who are interested what you were just talking about, right?  That was some time in the mid-1990’s or the early 1990’s?  

D:  Let’s get to there in a minute.  It’s a new concept.  It’s like saying that the payoff that seems to be needed is that people start collectively sharing in their development of their “C”, their knowledge, etc., of how to improve their organizations.  There is a lot, so much happening, but how can you depend upon a consultant?  It’s sort of like, “Andersen (or name one of the others), can we come and see how you’ve changed your way of organizing and working collectively in your knowledge, etc., as sort of an example?  Before we hire a consultant to come in we want to see how you’re doing your work, and how futuristic, and what your understanding of the future is.  How are you employing it?”  And with that kind of a test I bet wouldn’t be thrilled.  That’s what I learn when I talk to big consultancies, saying, “Well, as a matter of fact, our specialists sit out there and they guard very carefully what they know, because that’s the leverage that they have in their consulting business among their peers inside the big organization.  So how do I get the idea that the things you’re coming and telling us the things we ought to do really embody a picture about the future that’s really happening?  “Oh, you don’t trust us?  We’re big, we’re…”  “I know, but why isn’t it exemplified in how you do your work?  Your work is purely knowledge work out there where your company works.”

H:  How do you induce people to share knowledge when sequestering knowledge is a source of power?

D:   Well, you start looking at it.  “If you guys contribute money and people to this collective process of developing a collective picture about the future wave in which you could knowledge work and about the ways in which you could improve your own improvement process, you’d get so much more return collectively than you do if each of you is expending your x percentage of your budget on just trying to do that yourself, going out and coming in and describing the future as you see it, etc., and the plans and stuff.”  So it’s just a very simple picture about where the most payoff is.  You don’t have to tell them about the explicits of your redesigning your organization, but you’re talking about how you improve your improvements.

H:  So, how did that go, trying to put that consortium together?

D:  Slow.  We’ve had a number of almost-launches and collapses.  Now it  seems to be, the climate is steadily changing, and getting more so you can talk to people about this.  

H:  Well, at a certain point you moved into an office at Logitech, and that was a major milestone in support, wouldn’t you say?  When did that happen?

D:  Yeah, but that didn’t have anything to do with our big picture, it just is their showing gratitude for the mouse invention!  (laughs)  And that’s been a huge help to us.

H:  Well, what have been the milestones recently in the Bootstrap Institute beginning to get some traction?  

D:  The World Wide Web made a huge difference in perception, so soon after that, a year or so after that, we got enough interest to actually establish the Bootstrap Alliance, which is the not-for-profit framework for doing this.  And in our conceptual framework [are] these improvement communities, so we need to mention that a little bit.  Sorry to get… 

[end of the first side of the tape]

H:  Okay, so we’re talking about improvement communities.

D:  It’s like a number of organizations.  Maybe they’re interested in knowledge management, and so they say, “Hey, you know, wouldn’t it be very handy if we collectively started putting in some dollars and people into building up an interesting collection of knowledge about what’s going on today in knowledge management?  Instead of our going out there, it’s happened so fast, it’s just going to meetings and stuff …, so why don’t we do that?  As a matter of fact, when we’re building this, why don’t we try to use the best forward advanced knowledge management capabilities we can in order to build our repository together about the state and future of knowledge management as it applies to what we could do to change ours?  And what about case studies of different organizations who have been investing in changing their knowledge management practices, and such?”

In this whole thing, [???] people are contributing into this and are getting experience in a prototypical way of doing knowledge management.  “Great, then they can come back into our organizations to start showing us about how we can start changing, showing us the case studies that are out there that they helped collect.”  That’s the sort of thing, we call that an improvement community.  If they really do it in a networked way, then we call it a network community, because there are already scads of improvement communities.  So they say, “Oh!  Let’s just start encouraging organizations that want to invest in network improvement, and here’s how to do it.”  So our Alliance is helping people learn how to be better at mix at their common vector capability they’re trying to pursue, how can they better their mix.  And then we start thinking about the whole thing of improving infrastructure in society.  Look at the current infrastructure in any given industry, or something.  There’s a whole bunch of just traditional stuff—you have conferences and reviews and papers and talks, etc., and those things are exploding so fast, how much, how effective is that compared to what the new way of thinking and working ought to be?  So, anyway, it’s just strategically very important for society to start considering its improvement in infrastructure, and what could be done to strengthen and improve it. 

So we say, “I’ve got a place to invest advanced CODIAK capabilities, to learn how to do them, to boost the CODIAK capabilities that you’ve implanted in your improvement infrastructure.

H:  So what’s the state of that effort at this point?

D:  We’ve come up with a lot of good words.  

H:  You have some organizations who are committing themselves to at least talk about doing this?

D:  Right. 

H:  Do you want to name them?  

D:  That’s the awkward part, because in some cases there are people in there that are putting in resources and effort and it’s not formally “their company is in” yet, so they’re someone who can put in dollars or this or that kind of resource just out of their own budgets without having to go upstairs and claim it.  But it’s getting closer.  

H:  Generically, could you say there is a company that manufactures workstations, or a company that manufactures aircraft, or just to give an idea of what that kind of consortium consists of now?  You don’t have to name [them].

D:  There really isn’t a coherent consortium in place.  

H:  So you’re still building that.

D:  Right.  In ’91 or sometime back in there we actually had a number of corporations that got terribly turned on by it and there are people in them… this one company tossed in $300,000 just to help us with the cultivation process.  So we met every month or so for a year, people flying in on long weekends.  We had a very coherent plan, very detailed, it got ratified in January, I guess ’91, ’92?  I remember they had their corporate vice presidents all conditioned to sign off on it, it was very exciting!  But two weeks later the Gulf War exploded, and every one of those companies went into some kind of crisis mode, in which these special things would be moved up to some guy up above who’d have to approve of them, and who didn’t know anything about it, was overloaded, etc., anyway, and [there] was just no hope in cultivating them.  The whole thing just collapsed.  So that was the most dramatic of the almost-starts.

H:  What is the most exciting thing happening now?

D:  Let’s see.  That Peter and a woman both showed up in a year and volunteered, and they started really helping with the idea of professional development, what they call development activity in many of the not-for-profits, going out and trying to get [an] actual support.  So they started going out and knocking on doors and started getting responses.  Very interesting things happened.  So I really better not name names.  You can check back later before this book gets published.  What’s your deadline?  

H:  September 15.

Peter?:  You’ll have some names by then.

H:  Again, it’s okay to be generic.  It’s not as interesting, I think, to the readers of what the name of the company is but what kind of organizations are really beginning to be seriously interested in this.

D:  Well, Peter learned how to be politic about all of that.  He’s the one who has been generating a lot of this, so is this something you could say? 

Peter?  We could talk about the various projects that we are launching, that could be a stepping stone toward the far goal…

H:  Okay, just give me the high level description of those.

D:  One of them is, how do we … the ending of the book should describe this.  What’s a good way for trying to generate the content?  Someone came up with the idea of a seminar at Stanford, to do it on video, they have a video class set-up and probably doing that.  The video could be put on-line--now they can have it so it could be Web-accessible.  IBM people think that there’s a –name slipped! – a way to start addressing  video-frame links so that people on-line could make comments and questions about different frames and passages on the video.  

H:  That’s interesting.

D:  “Oh, well, while we’re at it, why don’t we just start making a dynamic knowledge repository associated with this seminar, and inviting people from wherever, and start coming and getting into a dialogue about it?”  Oh, well, while we’re at it…”  That would become a lot more important, the dynamics of whom we could get involved out there in the world than actually any set of students, even graduate students for this chart thing.  Okay, we already have provisions for having industrial people sign up for these seminar courses, so let’s start planning then to actually launch a PKR??? associating with supporting them.  And then in the floundering around, people would tell us that the faults??? in the quarter as good as winter quarter would be, to something like this.  So OK, we wait until Winter Quarter, oh, so we start getting people together and start talking about putting together this dynamic knowledge repository, and while they’re doing it, what kind of tool sets?  So why don’t we make the tools we put together those which have the best chance of evolving towards, and have a better OHS????  Then you start working on all that, and you’re really launching the package.  So the earlier packages we’d already trying to get people to do were less specific, but here let’s just get together and just start.  So we’re having a meeting in 2 weeks, getting people together, corporate people, other people who are involved.

Let’s see, the quite interesting part to talk about is…shall we tell him about the coming meeting?  Is that…?  I need an advisor here.

Peter?  I guess you know how it took a long time, not all the names, but check back on which names can be published.  I guess that’s easier.  

H:  Or can you describe them generically?  You can easily say “IBM” by saying, “a major manufacturer of computer hardware”.

D:  We have a couple of major manufacturers of computer stuff.  A very interesting national association in the education field.  A lot of interest but we can’t quite do the timing right with a major professional society.

Peter?  A major research company.

D:  Oh, yeah.  A major research outfit.  

H:  Okay, let me jump ahead here.  Disregarding how we get there, how would you envision a future?  Does it have to be 5 years, 10 years, or 20 years?  What do you think is the optimum way for systems, hardware and software and learning and organizational systems to evolve?  This is going to be a MIT Press book, they’re going to keep it in print, and in 10 years we’ll see what the future looks like.   We’re sitting in the future that we were talking about 15 years ago, and it’s pretty interesting how it evolved according to your original vision, and how it has fallen short of that.  So what I’m asking now…

D:  That’s a nice way to say it, rather than say that it’s different.  Go ahead.

H:  So how would you know look ahead?

D:  Well, I guess the most general thing for me is that we’ve really gotten the idea of the improvement infrastructure in society, and some of the key things to improve.  We’ve found that we’ve really got people thinking about this Bootstrap strategy in that respect, so that we’ve really, really been very busy improving the kinds of capabilities that help improve the improvement infrastructure.  Then when that gets rolling and there are a lot of things we could look at in society and in the world that need improving.  That’s our way to accommodate our environment, the way to accommodate our population, the “have-not” discrimination, corruption, just any number of things that we as a society need to find better collective approaches.  How to do governments, just loaded with issues in society that we have to find better ways to handle, and have to handle collectively.  

H:  Would it be fair to say this?  You really had to create the conceptual framework in the 1960’s and you called it “augmenting human intellect”.  A lot of that involved building hardware and software for humans to use.  Now it sounds like you’re talking about “augmenting social intellect”.

D:  Remember the language artifacts methodology?

H:  Yes.

D:  So the language and methodology…

H:  And training.

D:  …extended into the human systems side.  So that’s just been there all the time and it’s just had clarification in the last 10 years.

H:  You would see that happening by major organizations beginning to add another organizational component in the way that at some point they added human resources, and at some point they added knowledge management, that they would add improvement management?

D:  Sure.  It’s been there all the time because they do invest in improving, but it’s just sort of organic and doesn’t get the attention that everything else does, so they say, “Okay, there’s been an improvement infrastructure all the time.  A lot of it was implicit or unrecognized, haphazard or something, or the fact that we have to evolve a lot more rapidly just to keep from getting blown apart, sort of, with trauma.  So much is happening, and so different.  I was listening to PBS, Public Radio…

H:  NPR.

D:  NPR, and some place [in] the last month, Monsanto’s move into genetically changed material, that stuff has required something, there were some people raising an alarm.  There were farmers saying, “Well, hell, we’ve been organic farmers for the last 6-8 years, and proud of it, going like that, but if our neighbor next door starts planting some of that genetically modified stuff and the pods sweep over with the wind, and we get contaminated with it, we’ll no longer qualify as organic.”  So they just, the stressing in his voice was something like that.  And I thought, “Boy, there’s something where there has to be collectively some way of looking at that and handling that because of the advantage that the populace gets from genetically modified materials.”  How do you manage that?  So that’s an issue that there’s no way of telling how important it is.  It might be just of life and death importance to us, floundering into it without enough understanding of its consequences.

So how many things like that does it look like?  Every place you go you can find protest groups or radicals.  “Hey, get out of the way, you don’t understand business.”  But also the stable economies—how do we understand enough about the whole stock and financial system in the world to understand how to keep it stable.  And then things are happening, day trading and the like, which in a sense are what very rapid changes in the trading so that CEO’s of commercial companies have to be watching their standing in the stock market, a lot more closely than they watch the rate of change of long-term improvements in their company.  So the stock market doesn’t care whether you’re going to be there a year from now or something like that.  Gee, that’s an unhealthy situation in a marketplace, in which the people that own the company literally have no care about whether it’s going to be there in a year.  So it’s all over the place.

H:  So, you always have such confidence in reason.  These problems, if we are able to reason together about them, we might be able to solve them.

D:  It’s just my [???]  It’s true.  Then we have confidence.  I can’t think of any other thing to work on.  

H:  We do seem to lack a means of thinking about technology in the larger sense.  One way I talk about it, when I talk to groups of people, is that the original elevation of reason as a tool for solving problems, and the Enlightenment, really came about because so many… It was a situation in which there were a lot of plagues and wars and famines and a relatively small number of people said, “Maybe we just don’t know how to think about the physical world properly.  Let’s discard all this stuff from the Greeks and the Church and start from scratch.”  So you got Descartes and Bacon and Galileo, a relatively small number of people who said, “Let’s just start thinking about how we’re thinking about the material world.”  And in fact, they were very successful.  They came up with a set of tools, scientific methods foremost among them, for thinking about the physical world.

And that has worked so well of course that we’re gleaning problems out of our solutions.  What we need is another Enlightment in which we learn how to think about the world we’ve created.  We know how to create things, we don’t know how to think about what to do with the world, but it comes out of the ability to manipulate the genome, or to put significant computing power on everybody’s desktop, and connect them to the stock market.  All of the instrumentalities for doing things we’re very good at.  Figuring out ways to use those to better the lot of people who live in the world, that’s something we’re not very good at.  We have antiquated political processes for doing it.  So it seems to me that one of the underlying premises of what you’re saying really is, we need a new way of thinking.  It’s not just tools for solving problems, it’s a new way of thinking.  Would that be an accurate characterization?

D:  Well, yeah, thinking collectively.

H:  Thinking collectively.  

D:  Collective rationale.

H:  I don’t want to put words in your mouth here, but I’m trying to look at the way you’re looking at the future.  Would it be fair saying that in the 1960’s you were trying to get people to look at the idea that you could use technology like computers and televisions to do intellectual work and to communicate better.  And now looking at more explicitly, although it was implicit in that original vision, is, given our progress in technology, how do we think together about solving problems?

D:  I’ll actually go back to that 1951 high, and say, the world is getting more complex, the problems are more urgent, and they have to be dealt with collectively, and if we can’t boost our collective capability to deal with complex urgent problems, we’re going to be in trouble.  That’s what my commitment’s going to be.  It’s the same one, just getting clearer strategic thinking and ways to conceptualize it.

H:  There aren’t really very many people who have worked steadily on one thing for 50 years.  Know anybody else?  

D:  I never thought about it.

H:  Nobody comes to mind, actually.

D:  They’d think, “Geez, that guy must not have much imagination, he can’t think of more to do “

H:  I’m going to stop this here.