Douglas Engelbart, "Bootstrapping Our Collective IQ and the Aerospace Industry”

Search transcript...

ENGELBART: I had just left it set up, the laptop, and the projector was on too. Then I was sitting there and I noticed that I had not plugged in any power. So my batteries would probably run out halfway through the presentation. So when I came up here to find a way to plug it in, I inadvertently unplugged that. And so this would make a great start to an interesting environment between us here. And I'm going to have to reboot and--

AUDIENCE: It's working now, sir.

ENGELBART: Pardon? Oh? Oh! Thank you very much. That's the story of my life.

Okay, we'll be in there in a minute. One of the one of the things that caught my ear where Tom Ferguson talking about is the basic shift in paradigms that occurred in the launching of LAI, about the way in which organizations work and what you do your productivity, et cetera. And this is something that, by the early '60s, I had realized was a very big plus, that the paradigms that prevail with which to sort of guide the way people look at the world just have a surprising impact. And they've always been there. And there's a slow, basic change in paradigms that's sort of part of the evolution of a culture.

And it's very likely that it had a survival value in having that be relatively slow, so a segment of society didn't shift all of its beliefs and orientations too rapidly and fall apart. So perhaps that's the case. But it does take a long time for paradigms to shift.

Well, one of the things I really am coming to in this situation here is looking at the scale of change that the technology is catapulting us into, and that all of our paradigms are sort of very slowly emerging from the old days in which things changed relatively slowly. So how we and social bodies-- organizations, institutions-- how we cope with the future, you know, the real perspective of explosive changes in technologies, is sort of out of date. And that's a very basic thing.

And I learned it in spades by being fired and things like that. And applying for universities and corporations to come and get started in these things in times in which they just clearly said, well, no-- one very prominent computer-oriented industry actually turned me down in 1957 when I said I'm assuming you're going to go into digital work with computers. No, Doug, sorry. Not a chance.

So these really, really brought home to me this question of it. And so, if there were any kind of thing that could get across here from my perspective, it would be how do you get your perspectives and your paradigms to start readapting, really reassess the way they are? So one of them here is in time, the external environment-- which in this case, a lot of it is spurred by the digital technology and nanotechnologies that are emerging-- in which, over and over and over again, we sort of feel that's come here. And yes, there are a few adaptations and such. So our job is to take what the vendors offer and learn how to adapt it. It's sort of like that. But the changes that are available and the potentials are just way over that, and their impact on organizations is a really central, central issue in change.

And the organization's evolution in adapting to that stuff is going in these kind of jerky steps, including what LAI has done like that. And so how do we shift and cope with that really, really significant change in there? And this is just one of the very basic things. And with it come real problems. So it's just inevitable that, when all that collides throughout the world in all sorts of aspects of its business, of its economies, and the defenses and all that, that they become really strange situations and very big, very big.

So that's one thing. They said, well, how are we going to cope with those? Well, we cope with those by collective action and thinking about it, adapting our ways of doing it such, and finding ways to solve problems like that. So there are opportunities with all of this, too. And one of the big, big opportunities that just I seized upon-- you won't believe it, but it was 1951. I was saying computers. And we could interact with them on displays and shifting. The way you can couple your head in brand new ways to knowledge, the way you can collaborate with others in there, is just a huge potential there for boosting our capability for effective collective action in there like that.

Well another thing that, as a few years went beyond that like that, it's also, say. We can plug that kind of new capability back into the improvement infrastructure and improve the way we improve, which, with the challenge of how rapidly things are changing and the complex need for that kind of improvement, looked like a sure winner. And I just used the term bootstrapping. So that was driving me through those early years.

And then I ran into this paradigm in the mid '70s that, on one hand, the artificial intelligence people were saying, oh look. The computers are going to get so smart that the humans sitting in front of them won't have to learn anything. The computer will adapt to that. And on the other hand, the people were talking about office automation. What we're going to do is automate the way we do work now. And the real user is the secretary. And those two things just pushed me right off the research map.

But fortunately, we had enough doing-- we were already supporting, mostly, DOD units over the ARPANET and learning how to use this. So we had a small business. And we got that to be sold. And I went to the commercial world with that, which was a really good lesson. It just clamped the research down because the people who know how to run businesses says, no. You don't get involved with the customer. They'll just want things.

So then another very fortunate thing is that McDonnell Douglas with, perhaps, a foolish sort of thought or something like that, had been successful in commercially getting access to its CAD systems. So it wanted to go into the information services business in a bigger way. And it started acquiring companies, including the one that had bought me from the research outfit. And so bingo. I was part of McDonnell Douglas, which gave me a new freedom to start going inside their design, development, manufacturing people without seeming like a vendor. I was a cousin. And that was a real thought.

So it just became very clear to me then that there was just an immense value, that what we were offering in the way of collaborative work, with hypermedia, that we could do quite a few more things then than you yet can do with the web stuff, because it's really based upon where you can go. So anyway, these are the things I'd like to talk about in this sort of order. And this was an experiment, to give you an order that wasn't just exactly linear.

So the first thing to talk about is what we call the co-evolution frontier. And it's like saying, well, where does organizational capability come from? Well, it's just based upon your basic genetic capabilities, you know? You've got sensory, perceptual, mental, cognitive, motor capabilities. And these are the things you exercise.

Well, yes. But if you take a somebody that's never grown up in any culture at all. They're totally helpless. They're about like this.

So he says, well, why don't you look at the kind of things that do help produce more capability for people? Well, you've got a whole long list of that, including all the technologies you can think of. And so I was making a list of that back in the early '60s just to get organized about how you really can amplify human capabilities.

And oh, my gosh. But always, I began to realize, oh, look. There are a few other things, too, that are there. So there was no name for that. So I'd been calling it the human system. And then it becomes pretty clear that there's a huge amount of what we call augmentation over here, as well as that and the tool system.

So the thing [INAUDIBLE] talks about an explosive technologies on this right side, and one looks at how things such as the plow or the weaving looms or the printing press came in on the right hand side. And the ensuing evolution on the left hand side in order to harness that created big changes in there. So that the scaling thing is becoming really important.

And I'd actually done a study of scaling stuff in the late '50s, knowing that digital technologies were going to have to get smaller and smaller and faster and cheaper, and that was going to come by learning how to build them much smaller. So I got a chance to do a few months of study about scaling. Because once I had worked at Ames Aeronautical lab, my first job out of school like that, and had known aeronautics people and talking about Reynolds numbers and things like that, and that you had to get the same Reynolds number in all the pressure, so that the scale model would give you relatively valuable data, so I learned a lot more about all of that.

And one of the real things that came out of that was, if you're used to building and working with devices at a given scale, and suddenly a change of scale in some of the important parameters, as it changes slowly, some of the quantitative characteristics will change. But past a certain point it gets qualitative, and things won't work anymore.

And so I would try to give lectures. And I would just tell people, well, suppose everything in this room, including all the people, just suddenly scaled up in size by a factor of 10 and have three dimensions, would you notice it? And so everybody says, well, let's see. It'd be the same angle subtended by anybody's-- they'd be 10 times as far away, but 10 times as big. So I probably wouldn't notice.

But then you start pointing out, how much more do you weigh? Well, 10 times 10 times 10. That's a thousand times as much. Oh.

How much stronger are you? I don't know. Well, somebody in there would know that the strength of your bones and muscles is really proportional to the cross-section area. So you're only a hundred times as strong, and your chair is only a hundred times as strong. So that means, unless your chair is designed with a safety factor of 10, you're doomed. So you fall on the floor. So your bones will break because they can't stand it. And your lungs are only a hundred times as large for a thousand times metabolism needs. So you'd be dead pretty soon.

But anyway, this surprising set of things, when you're not used to thinking about that scaling, so that really came home to me in that study. So I looked at this diagram, and I said, wow. With the scale of change that's erupting and forthcoming here with the nanotechnology and the digital stuff, that the changes over here that are going to be taking place, some force-- because as you adopt all this and this it gets all of the turmoil here. But some of the change is to take advantage of that. There are going to be a lot of surprises in that. And so that's what I started really to pursue, and especially in this language in the getting your concepts symbolized, externalized, manipulated in brand new ways, and realizing that the way we're used to things as they're done in print, that's just got very little to do about how effectively you're harnessing your basic mechanisms in there.

So you ought to be able to read much more flexibly, getting the views you want, moving around. Much leads on there. But pretty soon you're doing things that look hard to learn and it's unnatural to use. So the mantra at the time was easy to learn and natural to use. And that's been inhibiting our exploration of this just immensely.

And even in the the Defense Department, I would go into DARPA and try to talk about some of the things, and they'll say, oh, you gave us this great system that will help people in a crisis. But they have to be able to learn how to use it in 20 minutes. See? And he says, oh. Well, let's see.

You've got some other important missions during times like that, like perhaps interceptor planes, right? Yeah. So with an important mission like that, do you expect any Jeep driver to learn how to fly that airplane in 15 minutes? Well, no. But--

So anyway, the perspective about how important missions would be in handling complex knowledge work, it just says, hey, wait a minute. It will be very important to go after high performance stuff. And there's just a lot in this domain that we could talk about all day. But we're not going to.

And so, any capability your organization has depends on the co-evolution of that thing, which includes stuff in your world, but also even in athletics world or just general business operations. So there is this something that I did was say, oh-- very crudely, because each tool system has many dimensions to itself, and so does the human system, but if you just do it a little bit roughly like this, you'd say, I've got a human tool space. And one's organization would fit in there someplace. So what if we go around the world looking at different societies and see where they fit?

Some use more or less technology, and more advanced human system development like that. And we say, okay. There are distributions, right?

And someplace in there, there's the organization you belong to that you're probably pretty proud of. And you think, it's right out near the frontier. See? We can anticipate today these different changes in the technology and in the human systems stuff. And we're a little bit lagging in the human system. But boy, LAI is taking care of us, moving us into here, right? Well then, let's just say there is a distribution like that.

And then let's talk about the scaling business a little bit. It says, well, now wait a minute. What if we really think that technology is moving at a rate that's faster?

Well, our whole way of changing our organizations evolved over centuries and thinking about where these boundaries weren't that far away. But what if-- this is why the paradigm shifted in people's minds. And they realized, oh, wait a minute. Here we are with our distribution. And here's all that you can anticipate today and the technology, with commensurate changes in here. We suddenly have a new frontier. So we should have different paradigms about how we move our organizations into that.

And then somebody else will come and look at it, like this guy from California, and he'll say, well, why don't you just be a little bolder about that? And you says, hey, we haven't really seen what nanotechnology will do. And it's just, that scaling things down, it's just going to produce capacity and capabilities and bandwidth and everything else that we never dreamed of. And oh, I see. And that's something. How is that going to affect the world?

Well, okay. What about computers and things will get so small that you could surgically implant them and they can get the energy to drive them from your metabolism, and they could connect to your nerves, and they can also have sensor things that go beyond what your sensory capabilities can do? And all that connects in. And boy, you could become, what do they call them, cyborgs or something? That's really tremendous. But the scale of things is going to be able to do that. Just a couple examples out of the air here like that, see?

And then intelligent agents are really just inevitable, going to come in. And they're beginning to. And hey, how are you going to learn how often you've got a whole crew of slaves? It's not realistic to think of them just sort of bumping along and helping do a spell check or some things like that. They're going to be things that you can learn how to steer and guide and will interact with. And it will take skill and learning how to do that, but boost your power a lot.

So he says, oh, that frontier is more like this. The relative scale-- and none of our processes for social or business adaptation, they all were evolved with the frontier being very near like that. So it takes a real call to re-examine what's the processes? What are the processes by which our organizations can adapt and start harnessing this thing, and our societies learn how to cope with it? So we need to look at different processes for our evolving. And this has led me to that bootstrapping. And then it leads to things like, oh, let's re-examine the infrastructure by which our organizations improve and see if the improvement infrastructure can be improved, and then if you can bootstrap it.

So that's the message. And it turned out that, for the formulation we evolved, that one of the key elements in that improvement infrastructure is what we called an improvement community, something to have special focus on. And LAI is an improvement community.

And then it's saying that look, if we can bring the modernization work, especially in collective knowledge work, into appropriate improvement communities, this is going to give us that extra second degree acceleration or something like that. It gives extra strategic leverage in the improvement process. And so that's one of the reasons why, when I talked to Earl, I thought oh, boy. There's a high tech improvement community and it's based in MIT, which has a few people that are used to working with information. And it's a great model. So let's talk a little more about that.

So one of the strategic things that somehow there have to be is ways to establish outposts out in that frontier. And it isn't a matter of going out there and setting up a group of people that, by themselves, are just doing startling things. The way that that's going to show organizations how to go is if those are doing real work. And it'd be very handy if the real work they're doing is a visible part of the work the organizations are doing already. So that means you don't set them up in a laboratory in a glass bowl doing their special things. So there are a lot of thinking about that.

And how do you get the high performance groups that are going to be doing that? How are they composed? What are they going to do? So all of that is an important part of thinking this way.

So we've talked about this picture of the co-evolution frontier. And let's talk about, then, one of the very important improvement categories-- the strategic, in there-- as a collective intelligence. So let's say you have a bunch of people, and they're involved some organization in this mandala kind of thing in here with all of the cross connections implying that every member of this organization needs to have a channel for communicating with every other one. And so that's the way they work together.

Then it's obvious that if they're going to get collectively smarter, they have to manage the knowledge in there and, in the traditional sense, treating a document as a container with some authored knowledge in it like that, you said, well, it's going to be documents. And with the way the web has taught people like that, it's easy now to tell them that the web is a beautiful way to start managing their knowledge. And how do you go about, then, making them smarter? Well, that's something we could talk for hours and hours about, too.

But the experience and the motivations and the things we experimented with just really show up a lot. And so it's how do you really move beyond the knowledge work capabilities that the vendors can offer you, because the user organizations aren't yet proactively up ahead enough to say, I want to start learning about the really far out things, rather than the way it is now that what we're going to think about is what we can implement today or tomorrow? And we have the political problems of conflict inside our organization of people that don't want to have to change that much, or are suspicious, all kinds of what you end up realizing are very natural cultural impedances for that.

And so in the end, there's going to be a dynamic knowledge repository that's electronic, hyper-media linked, like that. And the ways for organizations to manage that, they want to be a lot more alert about the outside world. So they're going to have to be able to go out and actively ingest and assess and organize the intelligence about what the outside world is doing. And this social organism that you call an organization is going to be judged on its IQ a lot by, oh, how sensitive and perceptual is it about what's going on in the outside world that offers opportunities or threats? And when either of those occurs, how smart and clever are their plan for how to cope? And how well do they, then, marshal their resources and coordinate them in coping? And how quick are they to re-adapt and shift?

And you begin realizing the way the future is going to go, that the annual yearly budget thing is probably going to have to go by the boards and become a much more dynamic. And a lot of things are changing. The recorded dialogue and the knowledge product are just, they're a way to say, oh, any given time you really want all the plans, the designs, the current state of things, the current state of your understanding about the outside world, the current state of the scenarios that you're depending upon, is all sort of what you call knowledge products, available for people to use in there. And how they study it and how it's organized is a whole field and exploration itself.

And then at any given time, there's a lot of dialogue about it. And you're in the midst of a dialogue exercise right now, this morning. And recording that-- well, what I say and what I present here like that, the technology is already available for that to get nailed down and be on a web, and automatically transcribing my speech as much well enough at least so that you can scan for that. So you can go back and then find out what I said and look at it again, and say, oh yeah, that's interesting. And then the way you can make comments on it are the links that you'll be able to provide can point to a particular passage that I spoke-- and its associated diagram-- or run up and down that.

Or cleverly say, hey, I want to hear what he says and then show the three kind of slide pictures that he showed that went on that. And then I'd like, when you follow that link, to highlight something there because that's what I like to talk about. Well, that's the kind of linked things that were part of our plan. And the basic underlying standards for the extended markup language as such will provide things like that.

So then people make comments on this and questions, and the dialogue's comment on the comments grows. So these are all things that technology can offer today. But it's just a introductory example of this recorded dialogue is a dynamic part of integrating intelligence and arguing about what's there now, by which your knowledge product evolves and changes. So every organization that survives into very far in the next century it's going to have to start living like this.

So there are a few other things to think about in this and say, oh. You know, if I'm a division or a part of a bigger organization, my division or project or program is going to have its own knowledge repository that's got to be dynamic and totally relevant to what we're doing. And yet, the whole organization does, too. And so every one of these nodes around here has its own dynamic repository, as well as the whole thing. And so all of those are evolving concurrently, and yet need to stay in sync, that none of them is looking at a future of something that isn't in any way contrary to any of the others.

So it's that concurrency thing that really hits the demand. Yet you don't hear that very often. You know they have to have interoperability, but not scaling. So the challenge is large. And the challenge there is something that you don't just get by getting a new product out by a given vendor or something. It's got to be something that evolves in a very different way.

And so we ended up with this acronym CoDIAK, saying that describes a core set of capabilities in the future smart organization, that they concurrently develop, integrate, and apply knowledge. It's the best we could sort of do. So we use that term a lot, and say, how do you go about improving that? And you can just guess that every one, almost every objective you'd like to set up for what LAI can do, can do a lot better if the organizations were better at this. And a very important thing-- oh, LAI could operate a lot more effectively. So where should you put your initial exploratory energy about doing outposts that are trying to learn how to do this? Well, strategically speaking, LAI is the place to start investing in learning advanced how to do this.

So I got some great lessons interacting with the McDonnell Douglas people. So they were showing me that, well, so you have a couple thousand people working inside of the McDonnell plant for on a fighter plane. And look at what you have in these tiers of suppliers and the kind of interactions and design coordination that have to go on there. So there's a real example for concurrency and interoperability requirement. So these are things that you very rarely, very rarely hear about when vendors are selling you things. But this is the real marketplace.

So anyway, this open hyper-document system is something that evolved in our minds like 14 years ago at McDonnell Douglas about saying, you really, really have to come out with a standard, open, hyper-document system. And the HTML offered a beautiful way to get a standard way for publishing. And the XML is starting to advance it. And most of that is being done by vendors, the evolution of standards. And what is critically needed, really, is the proactive involvement of the end user organizations in the evolution of the standards for this thing. So these are sort of the properties that all these things.

And then it becomes pretty clear that there's a huge amount of what we call augmentation over here, as well as that in the tool system. So the thing then, if one talks about an explosive technologies on this right side, and one looks at how things such as the plow or the weaving looms or the printing press came in on the right hand side, and the ensuing evolution in the left hand side in order to harness that created big changes in there. So that if this scaling thing is becoming really important-- and I'd actually done a study of scaling stuff in the late '50s, knowing that digital technologies were going to have to get smaller and smaller and faster and cheaper, and that was going to come by learning how to build them much smaller.

So I got a chance to do a few months of study about scaling. Because once I had worked at Ames Aeronautical Lab, my first job out of school like that, and had known the aeronautics people and then talked about Reynolds numbers and things like that, and that you know that you had to get the same Reynolds number and all the pressure so that the scale model would give you relatively valuable data. And so I learned a lot more about all of that.

And one of the real things that came out of it was, if you're used to building and working with devices at a given scale, and suddenly the change of scale in some of the important parameters, as it changes slowly, some of the quantitative characteristics will change. But past a certain point, it gets qualitative, and things won't work anymore. And so I would try to give lectures. And I would just tell people, well, suppose everything in this room-- including all the people-- just suddenly scaled up in size by a factor of 10, and have three dimensions. Would you notice anything?

And so everybody says well, let's see. It'd be the same angle subtended by anybody's-- they'd be 10 times as far away but 10 times as big. I probably wouldn't notice.

But then you start pointing out, how much more do you weigh? Well, 10 times 10 times 10. That's a thousand times as much. Oh.

How much stronger are you? I don't know. Well, somebody in there would know that the strength of your bones and muscles is really proportional to the cross-section area. So you're only a hundred times as strong, and your chair's only a hundred times as strong. So that means, unless your chair was designed with a safety factor of 10, you're doomed. So you fall on the floor. So your bones will break, because they can't stand it.

And your lungs are only a hundred times as large for a thousand times metabolism needs. So you'll be dead pretty soon. But anyway, this surprising set of things when you're not used to thinking about that scaling, so that really came home to me, that study.

So I looked at this diagram, and I said, wow. With the scale of change that's erupting and forthcoming here with the nanotechnology and the digital stuff, that the changes over here that are going to be taking place, some force-- because as you adopt all and this it gets all the turmoil here. But some of the changes to take advantage of that, there are going to be a lot of surprises in that. And so that's what I started really to pursue, and especially in this language and in getting your concepts symbolized, externalized, manipulated in brand new ways, and realizing that the way we're used to things as they're done in print, that's just got very little to do about how effectively you're harnessing your basic mechanisms in there.

So you ought be able to read much more flexibly, getting the views you want, moving around. Much leads on there. But pretty soon you're doing things that look hard to learn and it's unnatural to use. So the mantra at the time was easy to learn and natural to use. And that's been inhibiting our exploration of this just immensely.

And even in the Defense Department, we'd go into DARPA and try to talk about some of the things, and they'll say, oh, you give us this great system that'll help people in crisis. But they have to be able to learn how to use it in 20 minutes. See?

And he says, oh well, let's see. You've got some other important missions during times like that, like perhaps interceptor planes, right? Yeah. So with an important mission like that, do you expect any Jeep driver to learn how to fly that airplane in 15 minutes? Well, no. But--

So anyway, the perspective about how important missions would be in handling complex knowledge work, it just says, hey, wait a minute. It will be very important to go after high performance stuff. And there's just a lot in this domain that we could talk about all day. But we're not going to.

And so any capability your organization has depends on the co-evolution of that thing, which includes stuff in your world, but also even in athletics world or just general business operations. So there is something that I did was say, oh-- very crudely, because each, the tool system has many dimensions to itself, and so does the human system. But if you just do it a little bit roughly like this, you say, I've got a human tool space. And one's organization would fit in there someplace. So what if we go around the world looking at different societies and see where they fit?

Some use more or less technology and more advanced human system development like that. And we say, okay, there are distributions, right? And someplace in there, there's the organization you belong to that you're probably pretty proud of. And you think it's right out near the frontier. See? We can anticipate today these different changes in the technology and in the human system stuff. And we're a little bit lagging in the human system, but boy, LAI's taking care of us, moving us into here.

Well then, let's just say there is a distribution like that. And then let's talk about the scaling business a little bit. It says, well, now wait a minute. What if we really think that technology is moving at a rate that's faster? Well, our whole way of changing our organizations evolved over centuries and thinking about where these boundaries weren't that far away. But what if this is why the paradigm shifted in people's minds, and they realized, oh wait a minute. Here we are with our distribution. And here's all that you can anticipate today in the technology, with commensurate changes in here? We suddenly have a new frontier.

So we should have different paradigms about how we move our organizations into that. And then somebody else will come and look at it, like this guy from California, and he'll say, well why don't you just be a little bolder about that, then? And you says, hey, we haven't really seen what nanotechnology will do. And it's just that scaling things down, it's just going to produce capacity and capabilities and bandwidth and everything else that we never dreamed of.

And oh, I see. And that's something. How is that going to affect the world?

Well, okay. What about computers and things will get so small that you could surgically implant them and they could get the energy to drive them from your metabolism, and they could connect to your nerves, and they could also have sensor things that go beyond what your sensory capabilities can do, and all that connects in? And boy, you could become, what do they call them, cyborgs or something? That's really tremendous. But the scale of things is going to be able to do that. Just a couple examples out of the air, and you're like that, see?

And then intelligent agents are really just inevitable, going to come in. And they're beginning to. And hey, how are you going to learn how, if you've got a whole crew of slaves, it's not realistic to think of them just sort of bumping along and helping do a spell check or some things like that. They're going to be things that you can learn how to steer and guide and will interact with, and will take skill and learning how to do that, but boost your power a lot.

So he says, oh. You know, that frontier is more like this. The relative scale-- and none of our processes for social or business adaptation, they all were evolved with the frontier being very near like that. So it takes a real call to re-examine what's the processes? What are the processes by which our organizations can adapt and start harnessing this thing and our societies learn how to cope with it? So we need to look at different processes for our evolving. And this has led me to that bootstrapping.

And then it leads to things like oh, let's re-examine the infrastructure by which our organizations improve and see if the improvement infrastructure can be improved, and then if you can bootstrap it. So that's the message. And it turned out that, for the formulation we evolved, that one of the key elements in that improvement infrastructure is what we called an improvement community, something to have special focus on. And LAI is an improvement community.

And then it's saying that look, if we can bring the modernization work, especially in collective knowledge work, into appropriate improvement communities, this is going to give us an extra second degree acceleration or something like that. It gives extra strategic leverage in the improvement process. And so that's one of the reasons why, when I talked to Earl, I thought, oh boy. There's a high tech improvement community, and it's based in MIT, which has a few people that are used to working with information. And it's a great, great model. So let's talk a little more about that.

So one of the strategic things that somehow there have to be is ways to establish outposts out in that frontier. And it isn't a matter of going out there and setting up a group of people that, by themselves, are just doing startling things. The way that that's going to show organizations how to go is if those are doing real work. And it'd be very handy if the real work they're doing is a visible part of the work the organizations are doing already. So that means you don't set them up in a laboratory and glass bowl doing their special things. So there are a lot of thinking about that.

And how do you get the high performance groups that are going to be doing that? How are they composed? What are they going to do? So all of that is an important part of thinking this way. So we've talked about this picture of the co-evolution frontier.

And let's talk about, then, one of the very important improvement categories that's strategic in there, is the collective intelligence. So you said, you have a bunch of people, and they're involved some organization in this mandala kind of thing in here, with all of the cross connections implying that every member of this organization needs to have a channel for communicating with every other one. And so that's the way they work together.

Then it's obvious that if they're going to get collectively smarter, they have to manage the knowledge in there, and in the traditional sense, treat a document as a container with some offered knowledge in it like that. You said, well, it's going to be documents. And with the way the web has taught people like that, it's easy now to tell them that the web is a beautiful way to start managing their knowledge.

And how do you go, then, making them smarter? Well, that's something we could talk hours and hours about, too. But the experience and the motivations and the things we experimented with just really show up a lot. And so it's how do you really move beyond the knowledge work capabilities that the vendors can offer you, because the user organizations aren't yet proactively up ahead enough to say, I want to start learning about the really far out things, rather than the way it is now, that what we're going to think about is what we can implement today or tomorrow? And we have the political problems of conflict inside our organization of people that don't want to have to change that much or are suspicious, all kinds of what you end up realizing are very natural cultural impedances in that.

And so in the end, there's going to be a dynamic knowledge repository that's electronic hyper-media, linked like that. And the ways for organizations to manage that, they want to be a lot more alert about the outside world. So they're going to have to be able to go out and actively ingest and assess and organize intelligence about what the outside world is doing.

And this social organism that you call an organization is going to be judged on its IQ a lot by, oh, how sensitive and perceptual is it about what's going on in the outside world that offers opportunities or threats? And when either of those occurs, how smart and clever are their plan for how to cope? And how well do they, then, marshal their resources and coordinate them in coping? And how quick are they to adapt and shift? And you begin realizing the way the future is going to go, that the annual yearly budget thing is probably going to have to go by the boards and become a much more dynamic. And a lot of things are changing.

The recorded dialogue and the knowledge product are just there, a way to say, oh, any given time you really want all the plans, the designs, the current state of things, the current state of your understanding about the outside world, the current state of the scenarios that you're depending upon, is all sort of what you call knowledge products, available for people to use in there. And how they study it how it's organized is some whole field of exploration itself. And then at any given time, there's a lot of dialogue about it. And you're in the midst of a dialogue exercise right now, this morning.

And recording that-- well, what I say and what I present here like that, the technology is already available for that to get nailed down and be on a web, and automatically transcribing my speech as much, well enough at least so that you can scan for that. So you can go back and then find out what I said and look at it again, and say, oh yeah. That's interesting.

And then the way you can make comments on it are the links that you'll be able to provide can point to a particular passage that I spoke and its associated diagram, or run up and down that. Or cleverly say, hey, I want to hear what he says, and then show the three kind of slide pictures that he showed that went on that. And then I'd like, when you follow that link, to highlight something there, because that's what I'd like to talk about.

Well, that's the kind of link things that were part of our plan. And the basic underlying standards for the extended markup language and such will provide things like that. So then people make comments on this and questions, and the dialogue's comment on the comments grows. So these are all things that technology can offer today. But it's just a introductory example of this required dialogue is a dynamic part of integrating intelligence and arguing about what's there now, by which your knowledge product evolves and changes.

So every organization that survives into very far in the next century is going to have to start living like this. So there are a few other things that you think about in this and say, oh, if I'm a division or a part of a bigger organization, my division or project or program is going to have its own knowledge repository that's got to be dynamic and totally relevant to what we're doing. And yet, the whole organization does, too. And so every one of these nodes around here has its own dynamic repository, as well as the whole thing. And so all of those are evolving concurrently, and yet need to stay in sync, that none of them is looking at a future or something that is in any way contrary to any of the others.

So it's that concurrency thing that really hits the demand, yet you don't hear that very often. You know they have to have interoperability, but not scaling. So the challenge is large. And the challenge there is something that you don't just get by getting a new product out by a given vendor or something. It's got to be something that evolves in a very different way.

And so we ended up with this acronym CoDIAK, saying that that describes a core set of capabilities in the future smart organization, that they concurrently develop, integrate, and apply knowledge. It's the best we could sort of do. So we use that term a lot, and say, how do you go about improving that?

And you can just guess that every one, almost every objective you'd like to set up for what LAI can do, can do a lot better if the organizations were better at this. And a very important thing, oh, LAI could operate a lot more effectively. So where should you put your initial exploratory energy about doing outposts that are trying to learn how to do this? Oh, strategically speaking, LAI is the place to start investing in learning advanced how to do this.

So I got some great lessons interacting with the McDonnell Douglas people. So they were showing me that well, so you have a couple thousand people working inside a McDonnell plant for on a fighter plane. But look at what you have in these tiers of suppliers and the kind of interactions and design coordination that have to go on there. So there is a real example for concurrency and interoperability requirement. So these are things that you very rarely, very rarely hear about when vendors are selling you things. But this is the real marketplace.

So anyway, this open hyper-document system is something that evolved in our minds like 14 years ago at McDonnell Douglas, about saying you really, really have to come out with a standard, open, hyper-document system. And the HTML offered a beautiful way to get a standard way for publishing. And the XML is starting to advance it. But most of that is being done by vendors, the evolution of standards. And what is critically needed, really, is the proactive involvement of the end user organizations in the evolution of the standards for this thing. So these are sort of the properties that all these things--

My watch ran out of battery. So you can just tell how well I'm connected to today's world. So I had to borrow Earl's. So this means he can't tell me when my time is done. And that was a very clever tactical thing, don't you think? I should have borrowed everybody's watch there so he would have no way.

Anyway, here again, it's that sort of strategic need to get the end user organization, or the real customer is the organization, in the IT world that's coming up, and that's here now. So how do your organizations move into this space? And it's only going to happen if there becomes a real standard for the knowledge packages, that it's not a standard set by a vendor that this is my kind of file, or something.

So there's a standard file. And not only that, there's standard sets of functional capabilities that can be applied to those things. And with a standard file there, there are a whole array of as yet unexplored, underutilized properties that can be embedded inside the document.

And some that we built in ours, for instance, were-- it was just early experiments with it-- that the identity and the date, time, second stamp on every paragraph or line of code was the last edited time. And this was really interesting, the way you could employ that property. It's like, oh. There's some problems in-- our source code has all been hyperlinked from the late '60s on. So 500,000 lines of it, that you could just jump to any place. And our email was hyperlinked from 1970 on. So it could point to any place in the code or anything else, and you could also point to any passage in an email. And it was just living that way.

So there are new bugs in the software? Oh, well, they're probably in this section of software. Well, one of the things we said is every time you jump on something, your link can specify what view. And it can say, just show me the first line of every paragraph, or if it's hierarchically organized, just the first two levels or something like this. Or here's a content filter to put on. And only show me those paragraphs or lines that have a certain content or were changed within a given period.

So there's some new bug is in there like that? Well, just show me something that was changed in the last month. And you'll see five or three or four people that have been changing. And then one that always tickles me is, oh, Fred. He's always the problem. Just show me everything Fred has changed, and I'll find the bug.

And so we had all these kind of ways to live. And it's learning how to implant the kind of properties that give you extended capabilities beyond the space. And a document now is not a model of a paper document at all, with its rigid thing that's only you can scroll. You know, you can jump very flexibly, have all kinds of viewing options and things that you learn how to apply. So it's inevitable that that comes.

But that won't be interoperably available unless there gets to be a standard way in which the document is expressed. So there's a whole set of standards out there that are-- and the standards working group-- that are available for evolution. But it's just absolutely imperative that end user organizations get proactively involved. And one way to do it is through consortia like this. And we'll show you more like that.

So we actually have funding now to proceed on this. When I say we now, it's-- this whole picture about what's needed in proactive end user organizations emerged so strongly in the late '80s for me when I was at McDonnell Douglas. And not to point at them as anybody who was particularly reactive. Every organization had the same kind of reactions. But we demonstrated what you could do in things. And we're showing how you can make a link in the requirements or a project status report, and a link pointing to something in a CAD system that when it got to the CAD system, you got the protocols arranged, so it would evoke the right kind of macro that would produce a given view and scale and view of a certain thing and highlight it, so that that's the way that the user, in seeing this comment on it, could go look. Or in the CAD environment like that, they're talking about the plating or the finishing or something like that, and they could have in a comment sort of sense, a link that go phew, right to the actual specifications of those things. So all those things were just there to add to what we could do.

And it ended up in that the poor head of corporate CIO was really between a rock and a hard place, because he says, well Deck and Hewlett-Packard and IBM, none of them are talking about this. Course the language he used at the time was a little more colorful than that. But he says, I'm not going to go ahead and do something where out there, they don't even recognize, they don't even know what a link is.

So there's just a question there, like that the world was not ready. And you can't go point a finger at any particular person. The world just wasn't ready, for some reason that the evolution of the thinking in scenarios about the future, about the inevitable much more capacity and bandwidth per dollar that was just inevitable, that changing the organization is going to take a lot longer than the change in that technology.

So anyway one of my daughters who had been working with this area-- and also, she and her husband had gotten experience with a startup-- we set up this Bootstrap Institute 10 years ago to say, let's raise a flag and start telling the world. And it's been a very interesting experience. But the momentum is slowly picking up and accelerating now that there is a Bootstrap Alliance that's started, that says, hey. Let's get strategically set up so that people can participate in this.

And last November, by George, Japan set up a Japanese chapter of the Bootstrap Alliance, because one of the things we were talking about is, hey, this is a scalable thing. It's got to be, to have a infrastructure that really will work like this, it needs to be scalable, and to be scalable to national or global. And it'd be worthwhile to set up a national improvement infrastructure. Well, the Japanese aren't quite exactly saying that's what they're doing, but they're setting up their own chapter of that to be participating in that.

And so now we've come back home and say, well, let's get the US more proactive. And we're getting a few other leads from other countries. But the whole idea of really paying more attention to the improvement infrastructure and how it can be improved and the strategies for investing in it-- so anyway, collective IQ.

ABCs of improvement is the next one. And so this is condensing an abstracting quite a bit, but go look at any one of your business organizations in LAI. And just look at these three categories of activity. So don't say the categories of people or something at the outset.

But the A category is everything the way the business is done today, every operational activity that conducting in the business day, which includes hiring new people and training them to do the work that you're doing today. So then any activity that's improving the capability to do your A activity, just categorize that as a B. And very often, that's somebody that wears an A hat many hours a day, but on B, it goes into a committee that's talking about something changing or something, or planning, or actually implementing it, or some budgeted change project. And the change agents often spend most of their time in there.

Except then there's this other category that became important as we were thinking about it. It's what activity is there to improve the capability to do the B, to do the changing? And so in a real sense, all of you that are from organizations here today, have your C hat on. And we found a lot of very interesting things that start evolving when you're thinking this way, which again, you know, you could dig deeper into dialogue, and we could be here half a day or more on that. So this was very important.

And one of the big things is, when the frontier is exploding, B and C become a lot more important. And C, which hardly gets any recognition at all today in any corporation or something, becomes much more important. So let's talk about that a little bit.

It's sort of like, say, that all right. We've got a frontier out there that we're going to try to go after. And the C is sort of like the scout. It's sort of, in order to improve your ability to improve, which here would mean, in order to improve your ability to get out there and move your organization, you need to scout it, find out what's out there, and what would be the different trails that are available, and how you can sort of teach the B kinds. And you says, oh, the B is the guy that's got to-- he's the wagon train master or something. He's got to move you out there.

And so, in the business of talking to people about being change agents, both of them are change agents, B and C. But it becomes important to clarify when you're in a B role or a C role. And then it'd be very important for organizations to start trying to do their auditing and something so that they could tell you how much money they're spending, what the budgets are for those two, because just like budgets for product R and D or something, where you sort of, by the time you're setting your budgets, you sort of take stock of what the rest of your industry does like that, because if you're too far off from them, you're either behind the ball or you might be overspending or what. But anyway, this is something we haven't found any organization that even has an auditing set up where they can classify and tell you what that C is. And yet, the thing about this explosive frontier-- thank you.

So anyway, every one of these steps we're taking through here like that are things that there are a lot more digging into. One can get-- and what I would really secretly like is to get LAI as an organization interested in this whole strategy, and start knocking on the door of our alliance to say, hey. We'd like to participate, because the alliance is trying to-- its job is to help different improvement communities do their job better. So anyway, this is sort of the objective for organizations frontier perspective. And it's just a part of the improvement infrastructure we're talking about.

So the next step we can go to is talking about improvement communities. And here, you can almost use LAI group. So among your manufacturing organizations that are joining here, and even some of the agencies, you can sort of look up and down the street and pick out some of them that have a similar set of capabilities that they're interested in improving. And so in this case, the lean aerospace activity capability.

So you'd say, all right. We're doing that in common. So our Bs have similar kind of things to do. And sometimes we'd like to be a little bit proprietary about what you actually do in A. So some of the B things, we would just as soon keep on our belt. But wait a minute. How about the C?

Well, suddenly that frontier is going way out there. And we have to invest a lot more in C in order to be able to know how we're going to move our organization. So it really benefits us to invest collectively. And that's just basically why you're investing and participating in LAI. It's exploring and finding out what's out there to do and to change so that you can decide inside how to change and do it. So he says, really awkward.

So we just said, oh. So what LAI essentially is is a C community. You're getting together in a community sense to do common work on C. And so you say, that's really very important. So out in the world, there are a lot of consortia that are doing that, and other kind of improvement communities that don't call themselves consortia, or call themselves a professional society or a trade association. They're doing that. So these all have a special sort of potential in boosting the rest of society's capabilities. So this is just the description of it.

So one of the things we talk about is saying, oh. We talk about what if the improvement community decides it's going to really be a proactive evolutionary pusher towards having this collective IQ of itself a lot more effective? Not waiting till all these member organizations do, but saying, oh, if we started doing it ourselves, more aggressively, proactively, really learning how to do the collective knowledge work of the consortium, then we're doing a lot better job there, and the whole pursuit of how our members can improve will be facilitated and improved. But also, the very experience that they all can share by having this thing be as advanced as possible is a real enlightenment about what they can take back inside their organizations.

And very soon, in LAI, that would become one of its capabilities. It's interested in transferring into the organizations, too. Essentially, how you do this CoDIAK-- Concurrently Develop Integrate Apply Knowledge-- in a better, improved manner, would benefit every one of the organizations that's a member of that. So that could become a directed sort of pursuit of one of the capabilities, but also be a front runner. And it's not just the knowledge that that can collect and be in that dynamic knowledge repository. It's the experience because how things can work in there. And so that becomes very important.

So we talk about the improvement vector, the vector of capabilities that an improvement community is out after collectively improving. And then we talk about if they are using this kind of advanced stuff in a really proactive improving way, we've been calling them NICs-- Network Improvement Communities. So that an improvement community that's not actively doing that is just IC, and some of our members have fun talking about how they're just an IC, trying to be derogatory, when you really could become a NIC.

So anyway, this is the idea that LAI is a real candidate for that. And so Earl was telling me that plans that you've been working on anyway, and you've already got a website. But he says, oh, like extensions-- so these presentations going on-- and then the dialogue about them, and much more, can start being in there. And much more of the practice for how that begins to be an important part of the way you participate-- so it isn't just LAI central staff. It's LAI participants that get involved and have to expend energy in participating in that evolution.

And the way it works, and you says, well maybe LAI's benefits aren't going to be worth it. So you says, well, just look at what you're learning about how advanced knowledge work can be done. And that's a real pilot operation for what you could bring back inside your organization. And LAI's role can pretty soon be there to facilitate that, too. So it can be a multiple winner.

So anyway, this is the sort of the thing. So you consider the payoff to the aerospace industry if LAI converted itself into a NIC, and not just a web user, but an example of a distributed organization, actively and persistently improving its CoDIAK capability and collective IQ. So this is sort of a polite challenge, people. And LAI would become more effective, improving its members critical capabilities, and incidentally, early and direct boosting of the members' collective IQ for the [INAUDIBLE] like that.

So here's the idea for the Bootstrap Alliance, that's now going into its second year. It's saying, look, what are the member organizations in it? Oh, the idea is that they are NICs, and that the common improvement vector that this NIC is there to do is to improve the capability of its members to be better NICs. So that's the way it is. So there's some real potential in that because it isn't something you do just naturally. There's sort of no one product, no one team or something that's going to produce to the winning capability in there. It's got to be an evolution.

So every one of these NICs should be free to start learning how, doing it itself. And this is the place within the alliance to start assessing, analyzing. It's just like no one outfit in LAI is, in competing, going to design the end [INAUDIBLE] best lean system. There's a lot of evolution going on in how you get the best case studies and the assessment of it that you integrate in here so that everybody can watch. You make a better environment for all of the members to evolve in.

And so this is briefly about the improvement community. But then there is the strategy about how you can really get maximum leverage out of that in this bootstrapping sense. So that's sort of saying, if your Bootstrap Alliance is careful, that early on, it looks at what we call the bootstrapping index for our NIC. That is, how much are the capabilities that that NIC is trying to improve among its members are capabilities that, brought back into the alliance, will help be something you would distribute in every NIC would need? And so we just find some real value in that.

So you know, like LAI over here, it's not its dominant pursuit. But it's going to have good ideas to share. But there are some communities that we've been knocking on the door of that I think there is a series of examples we have next. So these candidates would have a [INAUDIBLE] bootstrapping. So the association of computing machinery has special interest groups in computer human interaction, in groupwear, and the web is hyper-media stuff. So talking those, which each is an improvement community, talking those into becoming NICs-- so we've already been knocking on their doors and saying, it's something where they should get funding for them to start turning into NICs so that they accelerate what they're doing and bring their own capabilities in an advanced way in for every NIC.

And there are a lot of things about organization and coordination that they don't foster very well. But good solid NICs LAI can become, would be able to contribute a lot about how you can better run and operate those. But these guys will produce much, complete a very lot of dialogue and studies. And so there are professional societies on strategic business intelligence, on intellectual capital, on scenario development, which is a very important thing. So their whole pursuit is how do you do better scenario development? So bringing that into every NIC would be very important for going into the future.

And it's clear that security is an important one inside of every organization, and workflow, electronic commerce. And there's no big problem in going out and finding improvement community domains, that if you could recruit them to become NICs as part of this alliance, that their activity would be very stimulating to every NIC and valued. So this is the sort of bootstrapping strategy like that, of how to do that.

And so this sort of finishes the picture about the bootstrapping. And there are a few other things to talk about [INAUDIBLE] saying. We have quite a few different organizations that we've been involved with talking. Some of them are getting quite active, and some of them are fairly passive, which is the way any consortia, I'm sure, starts like this. And internationally, the Japanese chapter formed in there, that we're in the process of forming explicitly a US chapter. And the current Bootstrap Alliance will be called the International, to try to help facilitate every nation, because every nation has a distinctive culture in which the way in there that you do the best at collective knowledge work would need to evolve and adapt to the culture. And you know the culture's going to have to shift and adapt to make that work.

And then at the same time, you have to have international collaboration and cooperation. So you have to have underlying standards that are the same. And there's a lot you can learn from each other. So it'll be interesting to see how the Japanese go after that. But one of the early NICs that they're really pursuing is trying to get end user organizations into this business of how do they start pushing proactively, which is a very important thing.

So this national improvement infrastructure, it's something that would be important. You could have a whole DOD improvement infrastructure if we wanted to, or an aerospace improvement infrastructure that can go deeper and broader than you guys go. And in the DOD, you look at the planning that starts at the strategic level with the scenarios and plans, and drops down into tactical or actual emergency situations, and you realize that there is a lot of concurrency issue about developing that knowledge, and that in a lot of ways, to make that organization a lot more effective, if they try to do it all by themselves, they're going to miss out on what the rest of the world and the commercial contributions and all the other organizations are doing to learn how to make organizational effectiveness a lot higher. So it behooves the Defense Department to participate in this.

So anyway, the sort of innocent assumption I have is-- and I've been innocent all these years-- that I say, well, look. There's something where there's a real argument that's been evolving for years and is honed and has withstood a lot of different sort of chopping attacks to say, oh, you don't know the marketplace and things like that. Anyway, there's something there that I think needs attention. And one of the best things that happened is for us to get actual dialogue with people like you in there. So it would take some specific commitment in LAI to sort of pull out some people and resources to actually participate in the alliance with the other emergent ones.

It would even be better, it would also be well, if they could go knock on other doors that they know to say, oh. This is a bigger scale thing. And you'll have to excuse my sniffling and my hacking. I came down with a mixture of flu and pneumonia a month ago. And boy, I have not recovered yet. It's amazing.

So anyway, this is what I'd sort of like to leave you with, is the-- my god. I just, I've got a set of pet peeves. And one of them is that the Apple Macintosh steadfastly only has one button on the mouse, see? And we had three. And somebody says how did you pick three? Well, that's only because for the current status, microswitches and stuff, three are all we could get on the mouse. And how you can work your hands better and connect better, it's something we just need to break away from. Anyway, that's a substance of another whole lecture.

Okay, so there's a whole open source pursuit. And I just have to close rapidly because, without Earl there, I'm trying to be awkwardly honest about-- uncharacteristically honest, whatever-- in closing on time. But there's this-- OHS is something that we're also trying to build communities that want to participate in its evolution. And every end user community like this should get involved in that, in what you're going to want to see happen. And [INAUDIBLE] Well hell, open source is a really brand new way to go after complex software. And high performance teams are a really important strategic part of all this thing.

And it doesn't mean you go to some university and say, make a high performance team that can show how dazzling stuff. It means recruiting and equipping and training a group of people for real application work inside your organizations. And there's some very clear ways in which, strategically, that can be done, and a very clear way in which that would facilitate evolution.

And that leads, then, to the sort of say, another thing that's necessary is to have in your tool system, have classes of user interface. So you could have sort of the pedestrian class, or even the handicapped class, whether it's handicapped this way or visually or whatever, but also grades of higher performance. All of it can work over the same knowledge base. And then it gives you the idea about high performance support teams as another whole thing. It would take a while to develop that in some substantial way in your minds. But that's a really important thing.

And a key utilization for them is to support larger conventionally capable communities, organizations or project teams. So that's another whole strategic thing that, if that got started, and introducing those into your world would add an element of-- it's like those outposts out in the frontier. It would out a very, very important element in a lot of ways to the way in which, not only LAI could evolve, but your organizations. So this multi-class user interface is very important. And our systems have had that in the past. There are many, many things.

So I think pretty soon I'm going to come to the end. A serious estimate was made that, in 30 years, the world's knowledge will double every three months. And so you say, wow. Just think. So assumedly, the world's knowledge about fluid dynamics, all sorts of things, the electronics, the many things will change, that are relevant to your world. So how do your particular knowledge specialists evolve to keep up with that?

What happens to universities? Who's going to provide the education and indoctrination that this really big changes? So anyway, how do you keep your knowledge repository dynamically updated? Is it every three months or every month or every week or every day? So it becomes a really, really critical thing.

So it's sort of like a human's reflexes. If you're in a complex environment like, say, you've got the ball and you're trying to do open field running. Boy, you've got to just pick up everything right away and integrate it into your new plans and changes. Well, more and more of the world's going to get like that.

Okay, I did it. I'm 42 seconds over.

MODERATOR: Thank you. I'm going to [INAUDIBLE] took my watch away. It seems to work pretty well. I know you probably all have a question or two for Doug, but we decided rather than having a question and answer period, we're going to have a panel later this morning. And I would encourage you, if you have a question, just write it on a piece of paper and leave it on the front table here. And Dan [? Ruess ?] is a fast reader. He can process knowledge quickly. And maybe he can figure out how to incorporate your curiosities into the panel.

I think we timed our break well. I feel like I have gotten a-- there's this cartoon I have which shows this odd kid with kind of eyes that are coming out the side of his head. And he raises his hand and says, "Teacher, teacher. Can I be excused? My mind is full."

Let's take a break. We're back at 10:30. And feel free to engage Doug one on one if you want to pursue anything further.