Bootstrap Institute logo Doug Engelbart's
   Colloquium at Stanford
An In-Depth Look at "The Unfinished Revolution"
Session 5
Bootstrapping continuous improvement
Douglas C. Engelbart1.*
- unedited transcript -

Welcome everybody to Session 5 of the unfinished revolution. You know a year ago they were talking about Engelbart's unfinished revolution; well my picture is the revolution is really just getting started. It is going to be mine, it's got to grow and become something that a lot of people take part in before it really gets going. All the years we worked on quite a bit of strategic framework for something that could handle the scale of that. I've been presenting things like that in the early part and we've got a forum going on line that's got a few people operating very eloquently. So we nabbed one of them to come in and start talking some at this session, it will service him right. Right? We're also going to try an experiment about getting more involvement with people here because one of the things that's difficult for me through all the years is I build a framework with different concepts fitting together. It's very hard to know if I just go through and describe them in some order like that. It doesn't seem quite to gel as a whole thing. So somehow we need to pick specifics and interact and kind of interlace those specifics with the general framework to try to help it grow.

One thing I badly need is, I need personally is, really more dialogue about these different aspects of this framework of concepts, etc. So were hoping that we start with something that's direct. So last week when we started opening up about what the open hyperdocument system could, there's a spark that got there party by Adam coming through with his very reserved, I'm kidding, his very enthusiastic sort of way, which is very good. We're going to spend a fair amount of time at the end of the meeting dwelling upon that and try to knit that into some of the larger framework issues.

We're going to look at the bootstrapping part of the
UnRev framework
  • We aim for a workable strategy by which all parts of human society could improve their collective capability to cope with complexity and urgency. 
  • We have already noted that finding a workable approach toward significant, pervasive improvement is itself a complex, large-scale problem.

The bootstrapping part of that is something in which people say, well how difficult is it? So you say, well if I have compound interest on an investment in the bank then I know that every time I get some interest it goes in there and then that generates interest. Though a compound returning is another way of looking at, some people call it double loop feedback, etc. but the basic thing is if you're going to make headway at improving things, improving capabilities of collective work and then you take notice that the improve in infrastructure of our society actually is generally a collective process. Well that's very good, that if we early try to emphasize boosting our collective ability, called collect IQ, it's a set of capabilities we've named CoDIAK. So if we say okay let's start working consciously to improve that and then as you improve it in clusters the effort around the world or should you deploy it best? Well that's a choice and it turns out if you want to look strategically one of the really optimal choices would be to employ those improvements you had within the improvement infrastructure. In fact the improvement infrastructure that's trying to improve your collective IQ and then as that gets better and better you start saying now I can start disseminating the capabilities out around the world more and a good place to do that would be in the other parts of the world's improvement infrastructure. So that's what the bootstrapping thing really talks about, and it's not all that complicated. This time we thought well let's take two kinds of things, one is a very broad scale issue and the other is a specific one and just look at both of them to see can they play together in a framework we're talking about? 

Orienting for huge ¬"solution scale¬"
  • Early introduction, Millennium Project's Global Challenges 
  • We'll persist in assuming that UnRev can only succeed if it applies to a truly scalable, strategic approach.

So this huge scale is one thing that needs to be entered and we spent time at the beginning and several people objected because we brought it up several times in different sessions about the global scale of the millennium projects and the fifteen big challenges to society that are judged to be very large. So you look and say, do those have to be dealt with collectively? Collectively means they're disseminated, that solutions are not going to be by one small pocket of people or something that have to be involved quite a few people in an assorted community that's going to be dealing on that. So those are the kind of scales that if the world can't produce the scale of ability and capability to do this collective IQ a lot better in scales like that you still going to be in a great deal of trouble, even though an individual company or group or team might be a lot more effective, it's just very important to get effective in the large scale and then you can easily point to things that we'll go do now. So this unfinished revolution I persist in assuming that it can only succeed if it applies to a truly scaleable strategic approach, scaleable. I'll keep poking at that and sometimes over the last years, decades, that this has gotten me in trouble with people because they'll say, well we want to get started and I'd say, well there various choices you can make and the directions your going to take, can't we talk about strategic framework in which to choose them, there just trying to get in the way. What we would like to do is finally get a broad enough spectrum of people over the different levels, etc. interacting and things so that something about this evolves as a strategic framework and assumingly it may be quite different from mine, but it might still be more successful in my picture but the idea of going after that is important.

We're lucky today to have a visit from Professor Ohashi from Japan. Turns out that he, besides being a very active guy in the worlds committees of various sorts of products, also was appointed to be the president of an organization that's called the bootstrap alliance Japan. This emerged after two different visits into Japan and so he just popped across the pacific like it was nothing to come here and maybe do something else.

Between us we worked out the idea if you talk about the quality movement in the world, that was something that's been under way for forty years, forty, fifty, and it just steadily assumed a bigger and bigger proportion, perspective on it's quality movement until almost every serious big manufacturing outfit in the world is involved in that kind of quality stuff. It's been a very successful thing about that, so we say okay in a way that's a model about how one got a whole bunch of stuff working in there. In another way it's also something that says okay what if we introduce, if were successful at boosting collective IQ in communities at large, how much can that start really sort of put new horsepower into the quality movement. So we say, okay that's a very interesting sort of challenge. We're not sure whither he's professor and he's used to giving lectures, so what he's going to say, but were following him then with somebody local who works at SRI that's been very involved in the quality thing to try also to point out some of the relationships. So are you ready Professor Ohashi to come and give them a talk? 

[>] Interlude
Quality and bootstrapping
Masakazu Ohashi, Dr. Eng., Professor, Chuo University, Faculty of Policy Studies. President, Bootstrap Alliance, Japan

Thank you very much. You sit here and think about with this is the scale there's a worldwide quality movement really working out there. On one thing it's like whatever else you talk about in the bootstrapping sense it needs to be able to integrate, there's so much value there and process, learning about the processes, etc. It would just be stupid to start setting off, we're going to have a separate way in putting together a human system processes idea that's where the CoDIAK's integration part of it comes out there, that we have to integrate, the integration of the worldwide. What's learned and valuable in that and what can we add it if you bring in the kind of collective IQ and improvement infrastructure enhancement and such we talk about is very important to do. The bootstrapping is just going to do it, it's going to have to merge and participate. That's why the collective work is so important that we can find ways to reach across into all what's going on there. Also the other aspect of it is a national program in there that every nation got to be able to participate, which puts a global stress. If your going to have an open hyperdocument system it has to support the collaborative activities throughout all the things he's mentioning but also the things we talk about and the interaction of that, it's just terribly important. So the next speaker, he's going to talk without slides. He's Norman McCracken from SRI, he's a specialist in quality etc, and you've got another title someplace in there too.

[>] Interlude
Quality and benchmarking
Norman McEachron, Ph.D., Senior Consultant, SRI Consulting

That's great. That's a more example of the kind of integration that you can say, well if everybody was interested in what I've been talking about just moves into that camp of total quality could we help them boost? You stop thinking well it's going to take a mixing in of communities with different kind of compositions from the ones that are generally total quality in order to sort of boost the kind of collective knowledge capability we're talking about. We ought to feel like there isn't anyway that you can set up a bootstrapping community which stands all by itself, it's got to be one that reaches out and really gets involved with lots of communities, and so there will be communities of communities, etc. This part of the strategic framework is how do you start cultivating what communities can be brought in early and what you can do with them? Today we're going to move on into the OHS kind of thing with some of the ways it can evolve. But when it evolves it really needs user communities because it's the interaction cycle with the user communities there the ones that have been focusing on the human system side of it to get the co-evolution you want, but also to get the interoperability you want. If you build an open hyperdocument prototypes and there's things that in the end couldn't evolve to fit these needs then we say back up, we're not going in the right direction. We can have fun making local pretty toys, but if it's a strategic push like that we need to keep thinking about requirements for growing up open hyperdocument system that there be in there. So this is the kind of thing we're going to be talking about today. I could of given you a long discourse and all kinds of aspects of bootstrapping, but I sort of tend to instead go through it in a cursory way and then get more examples coming out and like we point to later we'll see that.

Extending the framework with the critical ¬"Bootstrap¬"
  • 1. General--a model for any organization; 
  • 2. A large, proven example--TQM, an Improvement Infrastructure that was improved significantly and globally; 
  • 3. Specific examples--bootstrapping the OHS evolution process.

We want to talk something about the model for any organization and we'll get some dialogue going on that. We've already then heard about the TQM as a very significant big thing to do it. Then we want to do some specific examples later on bootstrapping the OHS evolution and the open source way of doing it. We are going to try opening up, in the middle of this thing, with more dialogue and since you guys are here we can grab you, you're going to be the ones elected to participate today. If people out at a distance, we wish we had environment set up so that we can really be doing real time interaction collaboration with them. So you've seen this diagram before. 

05-Eng01.jpgFig. 1

The thing about this capability infrastructure is just very real. One of things that would be very handy in my world is to start getting people interacting about capability infrastructures. What capabilities seem to come up to the top and to give an organization? Which subordinate capabilities down there do they depend upon? Which of those lower order capabilities are ones that many of the higher-level capabilities have to depend upon too as components of their infrastructure? So it's that sort of not a hierarchy but a multiple interlacing on down of dependants on there that gives it an infrastructure idea and so one of the powers that just came to me looking at that initially back in the early 60's was if we're talking about making a significant difference in how people can think and collectively work together, that's going to be applicable throughout these infrastructures. So as they are they'll change the nature of almost any component capability in there. Well when you change the nature of one of them down here, that'll make it a different candidate for inclusion in some of the higher level ones. Which independently, since they can do their knowledge work better, would probably change but now they can change about the way in which they knit together other things down there. So that whole infrastructure is likely to start changing. 

The thing that's changeable clear down at the bottom level down there is how we learn how to harness people's native sensory, perceptual, motor and cognitive activities. Those things have all also been conditioned through the years to work with infrastructure as it was. The very language we've employed, for instance, we grew up thinking that that's the natural way humans interact, well it's the natural way in the environments that we've been involved with of course. But some of those environments are well what's the natural language to use? Well if your going into some stronger discipline or something you need to start learning a new vocabulary, a new verbs and nouns and adjectives in ways to work like this, yes, to get affective in there. Then we can say, well we can actually start modifying the actual physical interface with which we interact out in the world because now that doesn't have to be quite like this. The ways I can communicate directly with people could probably get a much wider band if I learned better ways to communicate and that's one of the reasons why last week I talked about Neil Scott's activity with the handicap. Where they're really learning different molds of using the kind of sensory and perceptual and motor activities that those people are left with. It's a very important thing to say; okay we sit here and think that because we are endowed with a balance set of capabilities, about sensory and motor capabilities, we're doing it the right way. Now there's just a lot of change potentially down there so if you start changing the interface here, with the kind of functional capability of all the tools and then the human system change, these infrastructures are in for just huge readjustments. Happen stance is they will cause a lot of strange things happening and it'll be very handy as if you can get a better perception down stream of what's coming in different places so organizations are more ready for suddenly realizing that some very dominate capability right here is going to be replaced by two or merge with another one and be much more effective. So those things need foresight ahead to think about in looking at and planning and collective capability about handling the transitions of that, etc. So those are the B parts you have to get ready for and see like that. So you say okay you got a picture here, so then one thing we like to do is just have some dialogue among you guys and Marcelo Hoffman is going to be our group recorder and he'll call on me once in a while for any arguments or something. We're just going to see if we get people talking and what I'd like to talk about is this model and the idea of calling it a social organism, you say okay what does it represent? On one hand it can represent a person and on the other hand it can represent a nation or something like that. So I'd like some dialogue because I don't know what's getting across. So Marcelo you start. It'll be interesting to get the ideas of people. Well what specific examples of types of social organism would interest you? Some of them you might say I don't think that model will work for those or will it work or what can we do? Just taking that. What can we say about if you pick different social organisms to talk about? What's the adequacy of their improvement infrastructure within that regular infrastructure?

Audience Participation
  • Specific examples of social organisms which can be modeled this way? 
  • What can we say about the adequacy of their Capability Improvement Infrastructures?

PETERSON: One social organism that actually has a capability improvement infrastructure is the large industries. Very large companies, fortune 500 companies, they often use management advisory firms like Ernst and Young as their capability improvement infrastructure. They can select among different companies, those different management advisory companies give best practices. So in a way they've outsourced it to a certain extent. Which actually makes some sense because these companies like Ernst and Young can go around and collect ideas from many different companies. So I think the remarkable thing here is not how well the bear dances but the bear dances at all. 

PETERSON: I'm not saying it's a great system but at least these companies have actually thought about how do we improve our capability infrastructure and actually spend real money to bring in new ideas? Where as I think a lot of social organisms have absolutely nothing in terms of a formal structure to try and do that. 

McEACHRON: I had a similar reaction in some ways, being involved as I have been for last 20 years in benchmarking, I've seen it used as a systematic improvement process at several different levels in an organization. There's the nature of learning good practices from other companies, which is in part a process of humility, in part it's a politic process. Benchmarking that is looking at what other companies have done is somewhat akin to a scientific experiment that says, yes this really is possible to do, no we don't have to continue to do it the way we've always done it something else is available. The difference between benchmarking and traditional management consulting advice is that a consultant can sometimes get perceived as a person who's speaking in a way not quite realistically or is that really possible. 

McEACHRON: But in doing a benchmark you're actually looking at operative practices that have been proven successful by companies who have a track record. That way you get over the argument, it can't be done. You began to unfreeze the organization to admit the possibility of change. It becomes an objective, sort of a learning process, but a learning to learn process as well. 

ARMSTRONG: I guess I wanted to question the slide. I don't see that slide, in my view, as talking about a recursive process to improve anything but simply taken apart what exists. Mentally I construct the left hand side of that slide as the social systems, the communication systems, and the knowledge. Then there are the human systems that are actually working on human perception, learning, training skills. There's the tool set and then there's capabilities that we derive from the interaction of all those things. Any organization has those characteristics whether are not they're choosing to continuously improve them. Governments come to mind, just any social organization we have. 

EUGENE KIM: I actually had a very similar comment. I was going to say it in probably a more caustic manor. I'm not sure if I think the first question that you asked, what specific example of social organisms that can be modeled this way? I'm not sure if that's a fair question. I think that maybe a better question is, what social organisms cannot be modeled that way? I'm not sure if I can think of an example where you can't take any kind of organization and demonstrate how it relies on certain infrastructure both tool systems, human systems and so forth. The second question, I think is more interesting and I think it goes back to what your presentation was, actually I think, in the second session, one of the problems really is that it's easy to be sitting in a classroom and to say these are things that organizations have to think about.

KIM: I think it's easy for management to be sitting around saying we've got a deal with improving our infrastructure. The question is, what incentive is there for them to improve? I think the reality of the situation is a lot of time organizations are bottom line; they want to see if I'm going to invest in this improvement infrastructure I want to see profits or some other kind of bottom line improvement. 

RICHARD KARPINSKI: Well I'm interested in noticing what works and I've been in a field of computing where a very large fraction of the big projects actually fail. They fail typically after investing in enormous amount of resources that at the end it didn't come together. The one fellow that I've met who made it very clear how to completely avoid those giant failures is this fellow Tom Guilb and he had only three little details to take care of. One is, what is it your trying to do? What are your goals? To the point where you can get a numerical measure of how well you've met each goal. Secondly he addresses the problem of document after document being prepared with no check on whether we're still going in the same direction by doing some analysis in a very specific three person, four person group which he calls inspection. The final thing is never to allow more than about 1/20 of the projects resources to be consumed before you deliver an improved working system and if you do that then you can't have more than 5% failure at any one time.

KARPINSKI: It vastly reduces the problem and gives you redirection of your projects while your going along the course. It's like we do not start at one end of the Mississippi River and say now which direction is New Orleans? Now we'll set the boat and tie the rutter, no no, you want to make corrections very often, especially in big projects. So doing them in lots of little pieces is enormously important.

AUDIENCE: I was thinking of terms in a negative example that might spawn positive thinking and that is the way the software industry is tended to forget the lessons learned along the way. The specific example I was thinking of was speech recognition. Since 89 we've been using systems. We've had pigs and troths where when Windows came along all that we learned about DOS got thrown out but with it all the things we've learned about speech recognition got thrown out in terms of how it was used. So we started off with very primitive systems again and then as we went through each version it was obvious that a new team of people came on to create each version and they didn't go back and learn what we already new, and so we had these crappy designs each time that would then eventually evolve through and maybe catch up with the old ones. It just seems that we never seem to learn that lesson. 

AUDIENCE: I think we have examples of super organizational learning structures in markets where companies are competing for production of some good or service, like automobiles, or cameras, or stereo equipment. Because the metrics are well known and mature they all know that they're going to be competing, there going to be compared and they end up in kind of a coalition improvement infrastructure even though the specific improvements occur at the company level, there all sort of involved in a learning organism together. 

BYRON HALE: I've had the privilege of listening in to meetings of a large fortune 500 company, which I can't name but it's interesting to hear the management turnover, sometimes a new boss every 90 days. People who don't really know what there're doing and can't agree but seem to be determined that things are going to be done that way and that can take months or years to do anything about. Practically speaking you can talk about introducing all these ideas, but unless you have someway of goal setting that can somehow numerically evaluate goal setting itself I don't think that these things are ever going to be useful in practice because there to many egos evolved.

AUDIENCE: One thing that seems to being said is that the question should be not about social organisms but about social disorganizations. It seems that we often lump a group of people maybe at a company or a project or government whatever as an organism organization when really it isn't, it's just a group of people that happen to be in the same building or something.

AUDIENCE: I think we've been talking a great deal about organization and process but it strikes me that all of our symposiums, including the one today, one of the common factors in all of the things we discuss is values. The organizations are attempting to establish and produce values and I think that the complexity of trying to do that, whither it's with religion or products or whatever it might be, is at the heart of much of what we're here to discuss.

AUDIENCE: When I think personally, for instance, of a large complex problem, the kind we're describing here, I would describe it to myself as a problem that if you don't solve it, it will attack you. It will be a consequence for not solving an urgent complex problem and so often they are involved with again value. So the question arises, how do you deal with those things? How do you deal with the values that get in the way with the onset of the solution? As this gentleman here was just saying for example. 

AUDIENCE: So far seems to be most attention is focused on commercial type of companies. I think for government or military we probably need different kind of focus or different techniques. Probably the general framework can be modified little bit. From my experience with consulting to the military here, I think software project most company in the fortune 500 they can pay very high salary to very best programmer but for the military they don't have that kind of luxury so they usually pay very low and even they give a contract, doesn't give very high profit margin.

AUDIENCE: So I think we need to look at several factors. One is the incentive part of how can a long profit have a motivation to improve itself. Number two is the budget process, how can the budget process be improved? The third thing is the mentality of the origination, for example in the military most the commanders say, I want to conquer that hill in two days. They have the mentality but in software you cannot say, I finish this in two days. That's number one. Number two your military sometimes you can say, I want to send more people if I cannot conquer that hill I send ten times more people. But in software usually ten times more people gets you to cause more problems. So there are a lot of things like that so I think offering would probably lead to thinking for different organizations, different type of organizations may have different emphases 

BENZY DARA-ABRAMS: One of the things that I think is striking, especially in the valley, is the difference in teams working together in a start up venture versus in an established company. I think that would be useful in terms of looking at the infrastructure and how it operates differently in a high performance team that does get a lot of work done, a lot of innovated work and other teams that seemed to get bogged down in a lot of political discussions and meetings and really don't produce what they set out to produce. 

AUDIENCE: One of the common themes that seem to be underlining a lot of the scenarios we're bringing up. Someone mentioned the incentive that we have to change or improve. Well the opposite side of that is also what is the cost or the effort in time and dollars and personal career being at stake, for instance, in pursuing a particular path. I'm I as an engineer on the front line whose boss is saying, we need this result in two months, going to rely on my own wits or I'm I going to access somebody else's idea and give that idea a try? So there are some fundamental human issues, I think, in relation to this and part of what a knowledge management system, I think, needs to do is to help us access the knowledge that's out there with a minimum risk, maximum ease and convenience so that we can take advantage of it given our priorities. 

That was very interesting. A couple of comments on the outset that so many of the people initially assuming that this is a corporation. So that's fine, that's one kind of social organism and they're quite a few others in our society that are important too that we should improve. University or the whole higher education system, for instance, or the entire education system, for instance, in which each of those is a sub component capability. They improve the capability of the scientific teams around the world that together are trying to solve cancer problems. The people around the world that are trying, not only clarify, but tackle some of those grand challenges that the millennium project was done. Anyway, there're all these different kinds and in each one you can point to a lot of instances where today they are less than perfect. In a benchmarking sense, that Norman was talking about, when would be learn sort of go out and assess what the collective IQ of that organism seems to be? It would very nice if they had a benchmarking way in which there was enough collective interest in that in the world that just likes the quality movement they go around and can really assess and find people with leadership and give awards to the companies or outfits that do the best quality. How far away are we from learning enough about what the collective IQ's are to go out there and give awards to the companies? Another aspect to the thing is that's what consultants are for. They're the ones that are there to tell you where to go. A consultant comes and says, let me tell you how to do it, is one kind of help. That's maybe some of the work, etc, like that. How much difference that is from saying, we can show you, come and look at the way our organization runs. The quality movement has to do that where there is interaction between guys that operate drill, pressers, and shapers is also the engineers works, etc. So there is all these levels of interaction and they have to be talking to each other about this and the consultants can do help and guidance or something but if your really going to do an evolutionary thing in a big way, that seems to fall short. Unless the consultants are saying, we operate collectively with the highest IQ in the world, so you come and look and we'll help you learn. So if you find them like that that's great. It would be very interesting to start taking the leaders in the information technology business out there and say, would you mind opening the doors and showing us how you use information technology to boost your collective capabilities to do things? Don't just tell us about your products or give us a demo in a demo room about how you use your products, show us how you do your work differently and newly. That would be a big change in the world out there. So much from what I understood about the way the quality movement in the world works, it's sort of like that. The doors are sort of open for you to show from your factory what the quality activities you do. It doesn't mean you open the door for the product information or something like that that's in there, but these processes that make a difference in there. About nine years ago one division of a big company had just one this big prize, Baldridge award, they came to one of the meetings, our bootstrap meeting and they were all excited and they said, just think of this, there were these executives that used to be so stiff and closed like this, what are they saying now? They're saying, the thing to do is yes compete like hell out the front door, but cooperate closely out the back doors together. Things about that kind of attitude if the quality movement started getting that to prevail it's just a terrifically valuable human system cultural sort of improvement. The ideas of sharing, I'll talk a little more about that later too, this incentive sort of thing people are bottom line, this is talking cooperation. If the stock market keeps like it is now and the day traders are at work heavily like that, just every little dip a CEO is going to be
somebody compound on him for a dip or something like that. Much less for quarterly not meeting the good expectations, the stock price goes down a bit. So how in the world can a guy that's in that position invest very much beyond the next quarter? It's sort of hand tied. How did the board of directors deal with them like that? This whole perspective of the forced inattention of the commercial industrial, the people who have corporations on the market, forcing there to be so short term and so how does this speak about how long term things are dissolved? In the same way the market place, like in an automobile market place, once you got an environment where the consumers were busy doing their thing and there wasn't that much distinction among them, then the market place does a lot to drive up quality and price down, etc, like that. What do you do to help people and organizations boost their collective knowledge work capability? Which is going to be extremely critical in the future about how smart organizations can be, how well they harness their individual talents? Well if that can't blossom in ways because there aren't going to be sharing in the market place out there if the thing that the vendors of the information technology hype is so much short term in pushing and they're putting money into what they think they can sell and get market share. You look at that dynamic and you say, how much does that kind of dynamic serve this need for organizations to get collectively smarter? The only way I can see that the market place would work, in that respect, is if the organizations that want to get collectively smarter get a lot smarter about where they want to go and what the future really has. We sort of show this co-evolutionary frontier out there in a very simple two-dimensional way. There are so many paths out there that dig down deeper or take different roots, etc., that says, what's going to cost me to take a blind alley to listen to a given pitch from the vendors, etc., and go that direction? I don't have any other picture about where it's all going to happen, out in that foggy distance. So the thing about doing this collective work on scenarios out there and really keeping case studies out in the open in a study a lot of the things that the quality movement got, so that you can guide yourself, it just looks very important. Let's talk about schizoid organizations you see different factors in the organizations. Well you look at a real schizophrenic person which personality to have when you look at them in a given time, split personalities. You got organisms with example they're not going to be able to perform very well. What's going to happen out in the world of what a competition, or what ever it is, for organizations to realize that to reduce the amount of schizoid ness in them would be a huge difference? Not that they shouldn't have difference of ideas and opinions but if they have different commitments or something. There are a lot of things to talk about in here and I just very much enjoyed what you were talking about. Lets move and open up the possibilities of doing another one. Here's the kind of things starting to talk about bootstrapping.

05-Eng02.jpgFig. 2

Suppose I have a social organism whose major product out there is something that you know as a matter of fact is a product that can make a difference in how people's improvement, organizations improvement infrastructure works. If we sort of feed that back then we got bootstrapping. So the better you get at doing your product, the better your going to get at improving what you're trying to do which means this will get better which means you will get better at this. Now it's really that compounding thing. So that's the basic sort of picture about bootstrapping and it can work in any size infrastructure. So it's very important that we start learning how and different kinds of infrastructures you can introduce these kinds of capabilities to improve and get in there. This really means a lot, so that's the kind of thing to C community stuff is going to do. Well there's another thing to think about, what if it's an organization only part of its output is that applicable? Great, still worth bootstrapping, right? 

05-Eng03.jpgFig. 3

So different organizations or institutions out there the kind of major thing they're trying to do have different amounts of potential feedback. That means different amounts of potential for what they do can directly enhance the kind of collective IQ that can help the improvement and etc. Lets just generalize on that a little bit, lets say we have a number of social organisms SO, kind of cute number, so each of them has a varying amount of its output is valuable. 

05-Eng04.jpgFig. 4

But if we have something that's process of pooling what they're doing out there like that and feeding it back in to help their own capability improvement infrastructures, they'll all benefit by that pooling. This is really qualitatively different from saying we'll depend upon on a consultant running around here and coming back and telling them that it's qualitatively different. Consultant might work fine in some incidences like this but this is the way the quality movement did it and it just was extremely important for that. So this is what lead us to looking at this C community because a lot about what I just showed you here is really that business about what can help their capability improvement work better is the C. 

Capability improvement has to do it's own work is the B activity. So we're really pulling the C in there and that's what we show here in what we call a networked improvement community.

05-Eng05.jpgFig. 5

It's pulling the C and it's also then by being networked to say, lets really with this dynamic knowledge repository and better tools and methods for really doing that, we can really amplify a great deal what each of these organizations is going to invest in that kind of improvement thing. That's the sort of basic idea for the improvement community we talk about in accelerating and augmenting it. If you are going to go out early and you can only manage to recruit ten organizations to get in there to pull together, etc., so which ones would seem more attractive to you? Immediately you might say, some of them are trying to cure cancer or fix the environment would be very important for us to get, that end product would be important. Then the other thing to consider strategically, what are they doing producing if pooled back in would actually bootstrap you more, add more capability to improve and capability? 

Group discussion

Audience Participation
  • Can we list types and examples of social organisms which would seem especially attractive as candidates for early enlistment into a pooled CI-Resource cooperative? 
  • An evolutionary bootstrapping environment?

Lets just get into another phase of audience participation to deal with this. We have about ten minutes for that. 

HOFFMAN: We will like to do the same, get suggestions. Who would you suggest would be the best organization, organizations that will reflect facilitator? 

BENZY DARA-ABRAMS: One of the issues is who would want to put the energy into being an involved social organism and then the other on the flip side is who would we want? Those could be different because, I think, the ones that want to be involved are the ones that are having some difficulty or else their not going to feel that they need to improve. Those may or may not be the candidates that we want because we're looking for the ones that will have the high value to get back into the chain. So I think we have to look at that because if they're kind of fat and happy they might not think that they need to improve.

DARA-ABRAMS: In yet those organisms and organization might have more to offer.

ENGELBART: Sure, that's terrific. One of the things that any group that's going to try to make this work has to learn how to do is to develop the realistic value propositions to say, if you participate in this, this is going to be the value and to have to be strong enough to compete in all the other competing things you can spend extra resources on. So this is one of the things that the web has done a lot for is get people aware of, the change. We almost got people interested ten years ago but it's a very different world now. The business of a value proposition has to in one hand make relatively realistic the kind of pay offs from this and another one the needs, what's going to happen if they don't start investing in this? Another one maybe their pride about saying we're going to very early start working on this sort of assessment kind of thing to say, your corporation looks like it has a relative IQ about 40. Do you contest that or do you want to go. Let's shift a little bit. 

STEPHEN: It seems to me that the open source programming movement is one that's a great example where people of very large companies now, Sun, IBM, Apple are kind of rolling over and saying, this is the way I can get bug free code and it's good for all of us. 

ENGELBART: That's good because for the next hour and a half is all in that. Did anybody prime you for this? 

ANDREW PAM: Also I think there is a multiplier effect and you just have to be able to sell that. Basically once there is a momentum happening then anybody who comes in and contributes can get back more than they put in. Obviously if there's only two players then each of them is only getting back the same that they're putting in. Once you get to the point where there is ten or fifteen players then each of them is getting a huge multiple what their putting in. 

ENGELBART: That's a good point. Part of that value proposition has to sure them that those other guys are going to be honestly about contributing. So there is a way to design that to sort of help meet that. The other thing is when you get back, it isn't just knowledge like somebody can tell you about it, you have to have some of your people involved at good experience and that's how the design really works. Okay Norman.

NORMAN McEACHRON: Well I was thinking about joint ventures Silicon Valley and the other activities happening here in the valley socially and inter organization is a very interesting situation as well. There is so many clear problems in the valley that challenge the rich resources that are here and in fact some of those problems are barriers to growth in effect. Like traffic, which I feel everyday. So this is one of those areas were we could look not only at individual companies but also at inter company cooperation with multiple sectors for improvement in a fairly clear way. 

ENGELBART: That's a big part of why we talk about these improvement communities, etc., and call them NICs and then the levels that those can go to. Which you can have an improvement infrastructure that you can say industry wide or then state wide, so we're going to be knocking on California State one of these days and finding some progressive representative from that area that we can say, well what about it. So let's see, Eric.

ERIC ARMSTRONG: In terms of narrowing the search, it seems like the companies or organizations that are competing are probably least inclined to want to give their benefits to other organizations. 

AUDIENCE: Did that happen in the quality movement?

ARMSTRONG: I don't know, I see that sharing going on through organization, you have to correct me if I'm wrong, but through teaching but not from company to company directly. To say, I found a great way to improve our quality General Motors why don't you take this it's not going to be forwarded.

ENGELBART: It took another environment to think it collectively participant in to provide that. 

AUDIENCE: To get very concrete, I think, the main charity structure in Santa Clara County, which is where a lot of Silicon Valley is collapse recently, I guess it's being rebuilt. Also it turns out that Second Harvest America feeds approximately ten percent of the US population at some time during any give year. Those are two concrete examples. You might also think about why in a country that supposedly has such a great boom, that ten percent of the people find the need to be feed every year. 

ENGELBART: You look at it and say the market isn't working in that respect, and there are a lot of things the market place doesn't do, it's been doing some marvelous things but for a lot of collective problems like that there has to be some other way you collectively deal with it. We have to get collectively smarter at it. In one hand it's easy to ignore so people get all caught up in what their environment is and they can ignore that, but how human is that? That's another thing. Oh look what we've turned on; Australia gets a chance here because they say their going to escape tomorrow.

AUDIENCE: Actually those people who are on the verge of entering or just entering a particular social organization, are not necessary inure yet to the processes of that organization and there for are probably more open to seeing more than one set of processes and being able to compare and contrast and see where things are and aren't working. The problem is once you get inure to something it's very easy to not see even where it's not working. So I think narrowing down to those people can be a big help as well.

ENGELBART: That though has two sides to it, when you first get into an organization it's easy to see why it doesn't work but how you try to correct it knowing the personalities and the history, etc., is a second one like that. That takes a valuable thing. 

BENZY DARA-ABRAMS: Going back to the joint venture Silicon Valley example, per concrete example, out of that came Smart Valley and out of that came CommerseNet and CommerseNet is one consortium in which companies have worked together in this area. 

Would CommerseNet be a useful one to try to recruit?

DARA-ABRAMS: Yes, I think that's a situation where companies said, okay if we work together then we can over come some of the barriers in electronic commerce. They saw a reason to help each other. 

ENGELBART: We only have a few more minutes and let's try to concentrate on what kind of examples, and if you don't hit some I'll bring them up like that. Who has a good example? 

STEPHEN: When you think about communities of communities, the American Medical Association is one community of doctors but I'd like to see them benefit from being able to talk to the other communities of doctors around the world who are chiropractors or doing acupuncture in weird parts of the world or whatever. So they all kind of learn from each other. 

AUDIENCE: I'd like to just take the discussion a little bit to why we want to work with what company. It seems there are certainly lots of really important problems out there that need to be solved. But one of the main things it seems here is developing the tools, whither it's a knowledge repository or if it's an open hypertech system, whatever and I would think we want to choose companies to work with or organizations to work with that have problems that would be solved by the state of the art of the tools we have right now and where we can next take them. 

ENGELBART: That would be gratifying but will that be sort of lifting you up so you get better and better able to help them? 

AUDIENCE: Well it would certainly allow you to start working in developing the tools and get the tools up in a place where they might be used. 

ENGELBART: What we're talking about then is there are a lot of things that would be appealing, but just which ones that can be both appealing and their involvement would accelerate things? Yes, the bootstrapping retentive.

AUDIENCE: The exception of this room around the world there are some huge problems with aging communities and disabilities and one of the reasons that I decided to come to this colloquium, apart from wanting to come and hear more of Doug's work, was we want to see the growth of a global repository of knowledge that will help people remain independent. That I believe is a total candidate for the type of thinking we're doing. Where the person who knows the most about how to handle a disabled child in many cases the parent, and being able to share that knowledge with other parents who have just had a disabled child is a true bootstrap situation of being able to, we will help you, you put back the knowledge that's more fresh then we give you and we're bootstrapping our way.

ENGELBART: This would be appealing to me but I'll tell you the kind of things that I've been thinking about. 

If we involve the quality movement look at the kind of processes that they evolved in the past a lot of them would be applicable and all kinds of improvement infrastructure and bootstrapping. 

I think we're going to quite; we'll pick up on this after ten-minute break and everybody don't say a word during the break. We'll be back.

 - Break -

Welcome back to the second part of session 5 of our unfinished revolution discussion. We had a very good early part of this in the experiment about getting more interaction is very rewarding for me. I liked it. So in the second one we are going to pick up again about the OHS and focus for a while specifically on the open source approach with that. I'll give you some introduction and we'll get some speakers going. The coDIAK, which we remember, is concurrent development integration and application of knowledge,

Open-Source Pursuit of CoDIAK and OHS
  • Need participatory communities actively involved with using and improving both CoDIAK and OHS. 
  • Need initial framework for the hyperdocument architecture. 
  • Need initial framework for the functional tool systems, and Intellectual Property. 
  • Need basic organizational framework for coordination, governance, and evolution.

which is sort of the best summary and acronym fashionable thing we can get to say, these are the basic capabilities they're going to have to cultivate in order to get better at this collective IQ stuff. The open hyperdocument system we hypothesize is something that whatever the name is has to grow and involve in the world. No matter where you jump to with your links it's on a style that your browser knows how to deal with and it may have different grades of capability that you can click up to with different kinds of interfaces. You have to have the basics that your knowledge packages convey their knowledge in some standard ways and having structure in them, because a lot of our knowledge really is structured. Turns out it helps, in then having optional views, look at it in ways that for your particular needs at the moment helps you derive in what you need to know from looking at that knowledge container. We have known, by things that have been experimented with in the past, that there are a lot of very interesting things that people have done that aren't yet integrated out in the world. They have to be integrated and we know about that so we hypothesize that and it's a mixture of dealing with the content specification and the way the properties are encoded, etc, within the knowledge container. Having that be standard and at the same time they are going to be a rapid evolution, hopefully, and what kind of properties we find are really useful in there and that those get encoded in a standard way. Then the standard kind of operations you want to do in them need to be something if you are going to share views, etc, with people. Also their option for having wide ranges of capabilities still able to go in there with an appropriate interface setups with a lot more variety to it then today. Something that lets really high performance people move on ahead and function with a lot more capability inside the knowledge domains that they will be sharing with other people. We have talked about all these things in the past. This open hyperdocument system needs to evolve in order to fulfill the need for that kind of open, standard, global sort of way in which the knowledge containers are and the functionality for operating on them. The emergence a year and half or more ago of more and more awareness about the open source movement was just a very important sort of move for us to look at. Something like that would be the only way I could conceive that the evolution of this sort of thing we talk about in this open hyperdocument system. The only way that evolution can really happen in such a wide spent way, in which concurrently your going to have to be involving in many given knowledge domains but still in the way that you can inter-operate between them, etc, and also between different languages and things like that. So it's a dynamic way in what we're talking about. You need to have participatory communities that are actively involved in using and improving the both CoDIAK and OHS.

In order for having the evolution of the OHS be in harmony with the evolution of the CoDIAK capabilities themselves, you just have to have relative communities out there that are focused on improving the different aspects of it in cooperation. The functionality of that is something that the open-source mode just seems perfect for that. What we need from that is an initial framework for this hyperdocument architecture, an initial framework for the functional tool systems, and also a framework for dealing with intellectual property in that. We need a basic organizational framework for that evolution, you can't just turn it loose out there, and every open-source mode itself has got someplace that releases different new versions. Someplace there has to be some decision process that says okay that's version 17, we might have three versions 17 a, b and c, but in order to keep from having just random and staffing, you have to have some kind of process, whatever you call it governance or coordination in sort of dealing with that. That is similar to the process by which standards are set today that you have to have a community of people involved with setting the standards for any of these things so that has to happen, well those communities are all subject to being able to be improved in their capability by adding to their CoDIAK. We talked before the break about what kinds of social organisms would pay off the best to get started? If you got the kind of evolutionary communities that are involved with this stuff early on, be involved you see the experience they'd get would help guide them a great deal in what sort of things are necessary to build the CoDIAK and the OHS. There may be some of those committees out there that are so embedded in the old way of doing things and their all representatives that they might not want to move but it would be a real feather in the boot strapping cap to be able to make a value proposition to those kinds of communities that they get involved in the evolution of the OHS, CoDIAK too. They are getting more effective in their collective work, which would be a boost for it all. So I use the word governance in there and we need to talk about it right out in the open, it is like saying there's no way any collective activity can go ahead and get very completed without some process by which they decide which direction they are going to go collectively and how they spend their collective resources on this pursuit and what kind of standards they set up, etc. That's a real process that's not unlike any other governance process we see in society today. Governance processes themselves are capabilities, which are just marvelous ones that need improving, so lets start talking about that. I won't even embarrass California by saying, how far would you have to mature in some of these prototypical communities before you can reach out to a state and say, which state is going to get smarter first? Then you are going to start ringing the doorbell in different countries and say, why don't we start examining the improve in infrastructure that can be nation wide and which countries are really going to go about it? That's why we brought Japan over here, is to embarrass some of the other countries that might be tuning in on this and then you say, where in the country would you push the doorbell to say, hey wake up country, why don't you start getting serious about your improve in infrastructure and then help finding the ways you can improve your improvement process, etc. 

We will go from that back down a few levels to the open hyperdocument system and consider some of the things about it. The intellectual property that would start being generated is something that needs care and feeding. 

More about Intellectual Property!
  • We need to understand the implications of the various open-source licensing arrangements; 
  • And make our choice--for best chance that OHS and CoDIAK can evolve to the scale, capability and pervasiveness required for the challenge of adequately boosting mankind's Collective IQ. 

These questions have come up off and on through the years. They're the options to make it totally open or totally what. So that's the kind of thing that will get a very good treatment about here. So we have to learn more about it and then make our choice and for the best chance that OHS or CoDIAK can evolve to the scale, capability and pervasiveness that's required for the challenge of adequately boosting mankind's collective IQ. Think big. If you think big enough you find it just crunching you to the ground. I think its worth really thinking about so that's what were getting to do. We're now introducing Christine Petersen, executive director of Foresight Institute in Los Altos. I have known her know for a decade and I get more impressed all the time. 

Her husband is Eric Drekler who has done so much to get nano technology launched and I often think about if it weren't for Christine how big an impression her husband could of made on this, because she is very effective. She will tell you about more of the capability things that she's involved with. Christine.

[>] Interlude
Orientation: Open-Source Development
Christine Peterson, Executive Director, Foresight Institute, Los Altos, CA

Thank you, I'm a great admirer of the young lady here. Thank you. There are a lot of questions that still are there and so we know that there are different models and dialogue about the different models in particulars and if we get serious about going ahead with the OHS and the rest of the CoDIAK environment which kind of model would best pursue that. I don't feel comfortable with the benevolent dictator role exactly. 

I sit there and all kinds of things turn in my head but I think it would be better to go ahead with the sort of schedule we have. Some of the schedule things are there to fit, but I wonder if we have a little bit of unscheduled thing is Andrew, back here, is leaving for Australia and he was mentioning to me ahead of time that the things about the Xanadu project where there where lots of resources put in for years about that and some how it never wrapped up. He's got some things to say because he's been a close partner in that world for a while. So do you want to just speak from using that microphone there? Is that okay? 

ANDREW PAM: I think it probably would be better if I come up though because I will have to hold the button here for a while otherwise.

That's fine. Assuming it will take him a little while to get here, that this model that I showed some time ago is an old one but it embodies a conceptional thing that I think is a very important about getting ahead with what we want to have an open hyperdocument thing. One of them is different classes of users, another is different classes of devices out there that you are going to want to be able to operate from your cell phone or what ever else is out there like that.

05-Eng06.jpgFig. 6

So you need this sort of virtual terminal operator like that. So that whatever you're doing gets put into that. You also need to have the user profile and whatever different language and interface you're going to establish and stuff like that. It just has built into it a lot of what the potential evolutionary potential is out there that we've got. That's what I'd like to do and so Andrew tell them your name and address and what you want to say.

[>] Interlude Impromptu
Quality and bootstrapping
Andrew Pam, Australia

That does it, thank you. I didn't want to be impolite but some how what we need you to do is provide us with the links and you've in stimulated stuff ZigZag is a fascinating sort of gadget. Ted and staff have been working on it for years. It's the important integrated so the process is how do we; let's see, some place there was a statement about heretic rebel a thing to flout, my enemies came up and were shouting me out, but love and I had the wit to win we drew a circle and took them in. That's what this thing would have to learn how to do is draw the circles and start integrating and we got a purpose and the only way it works is by integrating lots of kinds of capabilities out there. So, thank you Andrew. 

Adam Cheyer  is going to come and talk for a few minutes about his stuff too and then we've got another speaker that's been preparing for talking about his thing. 

[>] Interlude
OHS: A proposal for moving forward
Adam Cheyer, Director of Research, VerticalNet, Inc., Palo Alto, Ca. 

This was Adam's slide and he didn't' sneak in her did he? No. So I think a good thing to do would be the next speaker here his Eric Armstrong and would I like to see is Eric come up here and he can comment some upon this and then he's taking off on his own slides and we can find those as soon as you've said something about this. All right? Meet Eric Armstrong. Do you want to just make sure that they see whom you are and you tell them a little bit? 

[>] Interlude
Further OHS-planning suggestions
Eric Armstrong, author, software developer, consultant. Under contract with Sun Microsystems

Thanks a lot. It's very stimulating to listen to that because I was just realizing with over thirty years ago we moved all our source code onto hyper linkage stuff and a lot of things he mentioned are things we learned we needed to do and found ways to do it. So it really kind of provides for me a challenge about how can we some times just sort of show the way a lot of those things worked together. As you start accruing them you find you have to start getting some generalized principals about how it's going to work so that each individual trick you can do is just a special case. Somehow we have to find a way to do that and you guys might all be saved here by technical difficulties. It's sort of by my inadequacy and moving around in this space. I can say okay you guys fix it. 

Thank you Eric, that was great. We ought to sometime describe more about all the things he's done, the books he's written, etc. So one of the things we know that a lot of communities that are going to be involved and the functions you have to do. 

OHS-Dev Sub-Communities
  •      Developing scenarios 
  •      Improving scenario-development capability 
  •      Collecting, analyzing intelligence 
  •      Improving intelligence gathering, analysis 
  •      Developing dialog conventions, practices, ¬.

So these are all basic kind of functions that in our DKR we've been isolating and listing. So the very intriguing thing to do is to do that sort of thing in the source code software development where we're trying to be as general and basic about the way the repositories work, about their back linking, about the attribution, about bringing in IBIS and its successor GIBIS, etc. in there. Then how to integrate the stuff that Xanadu offers and many other people have came up with them like that. But in a way in which there focused on there and the open-sourced thing where you saying, look let's really try to get in and augment, you know boost the capability the software guy is doing their work that one of the knowledge domains your interested in is the source code. So it should have it's special capabilities for embedding the properties you want but they still should be as general as you possibly can and so people can come in. So people that aren't all skilled in that can come in look with a simple form of an interface, sort of like that. 

05-Eng07.jpgFig. 7

Sort of a model like this, that you've got communities out there that are tying to boost their evolution capabilities and involving in that are all the methods, processes, procedures, customs, governance, techniques and everything else that evolutionary human system and they're busy at developing, integrating and applying the open-source tools. Which also involves the OH standard document structure, which we've been, assuming would be XML and that you get the communities dynamic knowledge base out of that. So this is just the model we would like to have in mind as we get all this started. So Eric's list of nice thoughts and such are very interesting, I get very intrigued in saying, oh this is how we do that with augment, etc. but augment its code structure and everything else is totally out of date but the capabilities in there are something to just show people. What I really like in some way in which we actually can demonstrate or else have some special sessions where we demonstrate and we don't try to web cast it or someway in which we do that. So anyway, there a lot of things I've learned about why we've got to have interoperability and I don't want to detail that to much.

Up front, about Standards:

Two areas in Tool-System development seem mutually and basically important here: 

  • 1. The unambiguous description, naming and usage of properties of our knowledge containers; and 
  • 2. The unambiguous description, naming and usage of the functions of out tools.

The standards for the unambiguous description, naming and use of the properties for our knowledge containers. So this is where the XML world and so we're going to get a real lecture of why John Bozack sometimes. Is that right? Wake up John and tell us. Then the unambiguous description also naming a usage of the functions of the tools, mistake, should be our tools. Anyway, those are all parts of what a DKR should really provide for us. So it's been an exciting sort of thing, I think, just a few little things to tailor with are the term. 

Vocabulary, world-wide ...
  • These aforementioned properties and functions are the nouns and verbs for what will be a very rapid expansion of world-wide knowledge-worker language; 
  • And it will be important to foster its effective evolution. 
  • We won't get maximum augmentation boost without cultivating a rich, powerful ¬"knowledge-workshop¬" vocabulary.

The vocabulary issue comes up for me all through the years and the developing and common evolving vocabulary is something society opens and society does and it sort of evolves. So this is what we need to do is the terms, the processes, the conventions people are going to use when they do some of this back link management and attribution support needs conventions. The customs, the terminology, the language we use, all that needs to evolve and there's just absolutely no way it can evolve with a proprietary push underneath it. So the open-source, open hyperdocument, open community for evolving the whole stuff is just what we have to have. So I think with that, even though I may have more slides here, I think it's time to close because I think our whole ninety minutes has run out. Thank you very much; we'll see you next week.




Above space serves to put hyperlinked targets at the top of the window
Copyright 2000 Bootstrap Institute. All rights reserved.