Doug Engelbart's    Colloquium at Stanford

An In-Depth Look at "The Unfinished Revolution"

I guess I was asked to get involved with this in some degree because we are either a guinea pig or a prototype and I'm not sure which. The Millennium Project is, as you've heard, is the beginning of the capacity of humans to think together in a cumulative way. Sort of what the doctor has been talking about. One of the mechanisms that we use for the - let's see, I should get his notes here and start learning his vocabulary now - the dynamic knowledge repository; we have a website; that is where you use your imagination. 

Imagine a grid. Going across the top are the domains. And we talked about having to have certain domains and standards on the domains. So we have certain domains for this: one is human capacity, another is technological capacity, another is environmental change and biodiversity - there are about five or six of these things. One of them is called integration whole futures. That's kind of like, you know, where things don't always fit in; you know, you stick them in right in that group. And then so imagine those 6 domains going across the screen. Now imagine going down on the left side of the screen are individuals, institutions, and maybe we'll have to add next, issues, opportunities, actions, challenges, websites, questions, and so forth. In this grid then, you'd click on one of these cells and it says what is the current state of play of actions to address that biodiversity issue. So there'll be a statement action number 1, number 2, 3 and so forth. With each action will be a range of judgments collected from one on one interviews around the world from policy makers who are supposed to have the responsibility to do that action because they might want to say that action won't work until we get a standard definition by scientists around the world. Until that time, we can't back that action because the business community won't back this policy because there is not an even playing field. Precise. Crisp. So we have a whole bunch of those comments, like a paragraph or 2 around each action. Now, where we're at, at this point is that we collect those information about the actions and descriptions of those problems through repeating questionnaires so that every year culminates into the next set of stuff and we produce an annual state of the book - that's right, annual state of the book is this. [Holds up book: 1999 State of the Future] Next year there may not be an annual state of the book; it might have to be the state of the CD-ROM and a better executive summary because it is very intimidating to walk into some financier's office and saying, "Read that, that's your homework for the cabinet meeting tomorrow." However, that becomes an excuse, a deadline for us to get stuff together; it's a mechanism for the management. We would like to move in the process of this to not only use the questionnaires to collect them in a batching way, and the interviews to get a judgment in a batching way, which is becoming rather large, and there is a lot of information in this system. And there's a lot for anybody to even get close to adjusting, even if they know what they're looking for. So we'd like to, in a sense, do some of what you [Engelbart] are talking about here. We've got enough up right now, so a person can go through each year: the questionnaires, who played, all the information is there. So you've got the grid, but it's also timed out, year by year. You can actually see how the accumulation of knowledge has gone. What I'd like to do in addition to what we're doing right now is to have anyone to click on, read it, and say, "That position is not correct today. There is now a new capacity that you don't know about." So there'll be an option - and hopefully within this year we might do it - a mailto function the person who clicks that on says here's how to change that. It doesn't go to a person; it goes to a preset list server within the millennium project to a set of identified and self-identified experts on that issue around the world because our stuff, unlike most of the stuff in the world today is trying to be inter-cultural, inter-institutional, and interdisciplinary. By inter-institutional I mean there's the private sector, the UN, there are businesses, there are corporations, government people, religious leaders, NGOs, so it's a cross-section of cultural orientations as well as the normal cultural sort - like the Europeans and Americans and that sort of stuff. So we would have a peer group, in a sense, set of individuals who that piece of email would go to. There would be a chair; that chair would have a month to collect the views from all the folks that may want to query that person back on something so they would have a sort of ad hoc working group on it. Then, when they come to a conclusion that that's the position - or they can't get to a position - then it once again, like the millennium prizes (?), says, all right, here's the range of the opinions on that. So you don't have to know the truth or anything about philosophy I learned as a kid; you don't have to know the truth but you've got to be clear about your confusion. So we would put that back up there so when you'd click back on that next time, it says, here's the other stuff and here's the evolution on that. So here's the dynamic on that; and here's our approach to that dynamic modeling so far. And one of the things that I'm interested about right here is how do we take the best that you all are discussing and try it out - this is where we become a guinea pig - in different cultures? Because we have nodes in places like Tehran; Tokyo; Magharai (sp?), India; Australia, about eleven different places around the world and each one is slightly different. They're supposed to be a little microcosm of the whole thing; they're part government, part UN, part corporation, part university and part NGO. So we're also trying to create a new organizational space from which new kinds of behaviors can emerge. They, then, cannot only be a source of new thoughts for the bootstrapping thing but also a place to try them out. 

Why don't I pause and take a couple of questions?

AUDIENCE: How do you know you've given all the voices a hearing.?

GLENN: We don't hear all the voices. We don't pretend to hear all the voices; we're not doing any kind of statistical polling. We're not here to say the average human thinks this. That's Gallup's job. We don't do that. If we wanted to do something about the future of physics in 1905 we might have interviewed two people and that would have been sufficient, maybe 3, all right? We're not looking for that. We're looking for what is the reasonable range of well-informed judgment. I work in a lot of many countries in the world and this philosophical fiction about making sure that you have this person represented and this person represented - well, this person has got his education from Stanford so what you're really hearing here is this Stanford professor's thing and not really India. So our approach to that is to create these cross-diverse nodes that are really different themselves so that the people they pick also allow them to be found. They then represent a range of views. Is this all the views that are on the earth? No. Do we want to have all the views on the earth? I'm not sure, [it] may be too much. But what I want is the range of views that have been vented by a range of people and scrubbed over. So it's not a Gallup and no, we don't have the ambition to do that. What we're trying to do is to help the human species think together about the future in a well informed way as efficiently as we can. The UN General Assembly is one approach - but one to argue about its efficiency. So no, we're not trying to collect all the views. If someone says that the earth is flat, then no, we're not going to write it in.

AUDIENCE: It seems to me that simply gathering an idea of what's best even is maybe not quite it. For example, look at sorting algorithms. There are many different sorting algorithms, each one essentially a deterministic output for many input variables and depending on which one is best for a given application depends on what the inputs are and what [are] the desired outputs. And so it seems to me that the best case would be to identify an unambiguous outcome of many inputs.

GLENN: Okay, one of the many ways that we're attempting to struggle with that is that we have an experimental node in Maui. At the Maui high performance computing center out there with some folks. And obviously we've got the project wrong; I'm in Washington and the computers are in Hawaii, but anyway. So you could see some weaknesses in the management right off the bat. [laughter] What we're looking at is creating a cyber game and if you can imagine a piece of paper folded in half. On this [front] side is a game, on this [back] side is the people with the day to day working as far as they're concerned; they're not playing a game but with real problems I n the world. On this side [front] in the middle we have all the knowledge we're accumulating and actions and various situational stuff and a person can come into here - and this is what we're going to design so if someone wants to design with us we'd be happy to do this - so someone comes into play this world game and they're trying to blow out this world problem, whether it is prectis [huh?] or war or whatever it is they're going to blow it out and they have certain strategies to do that. And they're get graded by some preset knowledge base over here on how well they're doing that. When they get to a point where you don't know - you're at the top of the thing in a particular area - you hook through to reality [front]. I.e., you win reality. So let's say you come up with a strategy to beat the virus in the AIDS thing that has not been thought of before and this little person here [back] - this is sort of like having their little expert groups - so these people have some sort of preprogrammed thing where they say, I'm working on AIDS, this is where we are, these are the things that we can't solve and this is where the next ideas are going to lead. You find someone like that in the world, I don't care who they are, but I want to talk to them, right then. So when that person gets to that point with a new strategy and there's nothing matching there, then boom - off this comes it's communicated to that person. Now why would these people play on this [back] side? Because if you had a world screening system to get to the person who had the knowledge - whether it's a taxicab driver or a president, it doesn't matter - then you might very well want to take the time to define your problem in such a way that is compatible to the game. Get the idea? So that's one way we want to - cause then we can do this universal knowledge stuff, this general stuff, but extremely precise at the same time through the game. So that's what the Maui team is experimenting around with but they're not a formal node yet but they're playing around with us; it'll take some time. And we were thinking of doing cooperation with them later because they're coming out with a new distribution system so we might want to have a game with they're distribution. You heard it here first.

But it's a few years down the road.

ERIC ARMSTRONG (PARTICIPANT): Do you have any thought to the kinds of features and functions you'd want to see in some dynamic knowledge system? Are you coming up with some predefined expectations of what you find to be good?

GLENN: Well obviously I didn't communicate it on the visual slide that was supposed to be the beginning of it. I guess I did need the slide after all. But that was the beginnings of it. One, you could go in and find out the evolution of anything, how it got there. You really wouldn't want to because there's just too much stuff, but one could and there would be the questionnaires, the rating systems, and who did the rating, and cross comment, and all of that up to a point. You could also see how that fits into the rest of the world situation, what's the relationship to it. We are also in the process of collecting indicators right now on how to measure progress on these questions and disaggregating that region: how do you measure that in Russia and how is that measured in Buenos Aires; that is going on right now. And we're going to interview about that - you know, how useful would that be for you, Mr. Chief of Strategy of Argentina, would you respond to that indicator or not, and why not? So we're in that now. That's in play and we want to stick that back and upgrade this grade, the stuff underneath the grid.

Now I'm very much listening where you would take that idea for a futures matrix. For those of you who have seen the movie [chuckle] . a little paranoia helps focus the mind. So we're calling it the futures matrix for right now. You can go to our website: www.acnu.org - 3 minutes left? Ok. What's the best 3-minute question?

AUDIENCE: There's a lot of controversy about the World Trade Organization (WTO) and the views that it holds. And I was interested that the domains that you listed actually didn't include economic activity and trade.

GLENN: I didn't finish them all, yes, and one of them is international trade and wealth.

AUDIENCE: Because this seems to be a great tool in fact to let all the voices be heard that are trying to be heard and maybe people should talk about it more as a place to put all of those voices. 

GLENN: Yeah. All is a big word. It is also a politeness. One of the things I learned from a boss of mine when I worked overseas in development and business was the right information, the appropriate next step - I'm talking about the appropriate technologies, also the appropriate information, appropriate knowledge, appropriate management - was the next piece. I'm really guilty of producing a book like this for crying out loud, and this stuff is really distilled down; the actual stuff behind it is like this [spreads out arms]. But I'm really glad - yes - economics and trade and wealth. And also wealth, we're also looking at the future of wealth because if we have wealth in the way we're going to have it today, we'll need three earths to maintain what we're doing. So we need to change the nature of wealth and that's very much in the dialogue as well. Yes?

ANDREW PAM: Yes, I just wanted to comment on the concept of hearing everybody's voices; that's a long standing threat in computer collaborative discussions: USENET, email mailing lists, more recently web discussions. More recently, the new news site slash dot [www.slashdot.org] has been looking at that where any time you allow a fully open discussion system, not all voices are constructive; you get flaming, you get people whose voice to be heard basically want to interfere with the discussion. So you can't really do that without filtering.

GLENN: That's right. Our strategy for that so far is that we have a list serve of people who are actually doing the work in the project and - forgive me
everyone who sees this [the video I think] - not everybody is there because they are more verbose than others; so even though they are regular participants they are not there. Sometimes we get our fingers burned and that is just the nature of progress. The other is we have a public one. Anybody can get into our website, self select; they can do whatever they want to do. Now we can delete them and they can re-sign, and delete them and re-sign, but that's how we manage that. Now, I saw a wave in the back: we're out of tape? Yes, okay. Yes?

AUDIENCE: Could you say anything of the effects you've seen so far; you've been gathering a lot of information for the past couple of years; you've been condensing it and putting it out there. Now, have you seen that influence any decisions being made or processes?

GLENN: Yeah, it's a real touchy business because - I'll give you an example; the UN University did a study several years ago about the use of Japanese surplus in economic development. The next year Japan's foreign assistance doubled. Now Japan does not want to say - this is getting out of line, isn't it? [Consults with someone] Oh GOOD, this was off the record. [laughter] Now Japan doesn't want to say that the UN influenced its decision. [laughter] You'll see that in Siberia next month. Decision makers don't want to be told; it's a gentle process. When we do our interviews one on one, it's also a debriefing for decision makers; it's part of the dynamic process. So not only are we saying, Mr. or Mrs. Decision maker, what's your view on this? But this is also a briefing. We've had UNEP do a bunch of stuff and development stuff and okay: one minute? Or your finger's cold? OK; one-minute question.

AUDIENCE: The difficulty with the inside and outside group is trying to extend the decision making process so there's only the outside group and during the filtering for serious content versus all the rest, have you touched any of that. 

GLENN: Yes, people can move from the quote, open system, into the professional system and the way that they do it is fill out the questionnaires and do some of the work. Now our work is not easy - and I can show it to you on one of our questionnaires and indicators - and you fill that out, you do that; and you have a right to walk through the door. So there is that system. Now at the same time I was a part of the New Jersey Institute of Technology's EYES (?) experimental study system back in the '70s or whenever, and when somebody didn't act in a nice way, we deleted them. The person barely talks to me to this day. But, you get your fingers burned in the name of progress.

AUDIENCE: But that's a good thing.

GLENN: What?

AUDIENCE: That they don't talk to you anymore. [laughter] 

GLENN: Okay, yes? All right. Sayonara. See you in the future.

[  principal lecture]

 ---