 Guten Tag, meine Damen und Herren. My gift to you is that the last German I will try to speak today. My name is Ian King. I'm from the University of Washington in Seattle, United States, and I'm here to talk about sort of a morphing topic. I'm going to start with one of the other hats I wear, and that's why many people here know me. And that is because I worked at the Living Computer Museum from 2008 through 2014, and in fact, I'm one of the people who created the museum. Myself and a gentleman named Richard Alderson were the spiritual parents of Living Computer Museum. I'll talk a little bit about how it evolved, how we created something from, well, a pile of junk, and how we had to work within certain constraints, and we had to be very pragmatic about how we approached having something open. This led to some interesting cogitation on my part, because I'm strange in that way, and so the talk is going to morph into a discussion of some preliminary results from my dissertation research, which is talking about a different kind of approach to understanding historical innovation. So, start talking about Living Computer Museum. So, when we first started to build it, we had some stuff, and one of the questions we were asking ourselves, not so much, what is a museum, but where do museums come from? How do you take artifacts and tell stories? And this was sort of a different approach to some other museums we'd seen, where they said, look, we have artifacts, and this is great to have artifacts, and museums have a very important preservation function. But, in our opinion, it was also about tying those artifacts into some kind of human stories, so that people understood why it mattered, or maybe even why it didn't. You'll find that I like to approach negative results as well. Negative results sometimes bound our positive results and clarify them, and so we were never afraid to take failure as a data point. We were never afraid to say, hmm, who isn't here? Who isn't connecting? So, Living Computer Museum started with really a pair of intertwined stories. One of them is the well-known story of Bill Gates and Paul Allen, who met at Lakeside School, the one sitting down is Paul. The really scrawny looking little kid there is Bill. And the parents of Lakeside decided that their children should be exposed to this new thing of computing, and so they signed up with the Genie Time Sharing Service, which turned out to be ruinously expensive and just going through their bake sale money. So, they managed to connect with a company that would lend them time on a PDP-10 that was primarily running the basic programming language. For one thing, this is the beginning of Bill Gates' fascination and ongoing love affair with basic. He has preserved it through many incarnations at Microsoft. The other thing is that it led to Bill and Paul learning more about this thing called the PDP-10. I'll just do that. So, the PDP-10 was Dex's answer to IBM. They really wanted a piece of IBM's lunch in the mainframe market. PDP-10 was the direct descendant of the PDP-6, which was actually the first computer commercially built to do time sharing. So, for people who all have their computers at home and they just connect over the network, time sharing doesn't sound like much. Time sharing was a critical innovation in that it allowed multiple people to use the same computer seemingly simultaneously. Many of the features that were developed on such machines, including messaging and email and communal work, later had to be adapted to network computing. And we see that with the introduction of workstations that then needed some kind of networking fabric. Now, the other part of this story, this entwined story, was Bill Gates and Paul Allen got connected with a company called Computer Center Corporation. They were selling time on a PDP-10 to all comers, anybody who needed some computer time, because most people couldn't afford to go out and buy a million-dollar computer. So, they would pay by the second for CPU time. Well, Bill and Paul made a deal with C-Cubed, as it was called. The operating system for the PDP-10 was still pretty buggy. So, they got C-Cubed to convince DEC that they would work on debugging the operating system. And as long as they were finding bugs, they had free accounts. Now, Kid in a Candy story doesn't begin to describe this. Paul Allen has told stories about dumpster diving. To get source code listings. And they were very successful at this. Well, they really saw the potential of computers and they thought that this new idea of the microcomputer was really interesting, but nobody had built a platform that they felt they could build any kind of system around. When the Altair 8080 was introduced in January 1975, I received that magazine in the mail, Paul Allen got really excited and said, here's our platform, let's do something with this. Well, at the time, this little guy over here sold with 256 bytes of memory. But it was also available in kit form for about $300. Prior to this, hobbyists had had to go around and get all the parts. So, they said, this is the perfect platform for things we want to do. What are we going to do with it? Well, it needs to have something that will support programming. Not everybody wants to learn machine code. So, they sat down and wrote a basic interpreter that ran in 4K words of memory. They didn't have an Altair 8800. So, what they did, Paul Allen wrote an emulator for the 8800 that ran on the PDP10. And that was how they created Altair Basic. About three years ago, Paul Allen reproduced that project because he still had his original paper tapes, and he was able to reproduce Altair Basic from the original code. So, that was the story we started with. And Paul had this fascination with PDP10s. He had a couple PDP10s. He had them available online through a service called PDPplanet.com. But he kind of got this itch to do more. And we talked about building a brick and mortar museum. So, in order to make a fuller story, we sat around and brainstormed about what do people want to see? Now, this is a fascinating story. The Bill and Paul story in Seattle, especially, is fascinating. But there are a lot of other stories. And, more importantly, what did people need to see? So, what were the things that we really wanted people to learn from the story of history about how we got to where we are? And what kinds of contributions could those make to their understanding of their lives? What are the stories that connect those things? So, again, you have a PDP10 here and a PDP7 here and a this and a that and the other thing. So, now, how are these at all related to each other? That was an important question we had to ask, too. How do we draw the lines? And that came down to also understanding who was reading these stories. And we thought about what kind of customers are we going to get? Is it just going to be people like us? Or are we going to see the general public? Are they actually going to have an interest in what we have to say? So, here's some photos from when we first started putting together the physical presence of the Living Computer Museum. This DEC PDP7 came from University of Oregon. It had been in service there for about 40 years. When my colleagues went in to see it, the professor who had been protecting it from being scrapped turned it on, became up and it ran. And he wanted to see it go someplace nice, someplace that would take care of it and continue to tell its story. And there were a few bumps along the road, such as the University administration, that wanted to charge us a third of a million dollars because even though the machine had originally been about $15,000, they thought, well, it's, you know, more valuable over time. We worked with some on that, got that down to a more reasonable price. So, we inherited this machine, very important machine. Unix was developed on the PDP7. Great story, great piece of history. We also had some VAX 11780s sitting around. Now, this is the first one that I restored, the 11780-5. It was actually upgraded 780 CPU. So, they had a logo for it that let everybody know you were the cool kid in town because you actually had the 785 CPU, which was substantially improved over the original. But this machine introduced in the late 70s really did end up taking a big bite out of the mainframe market. We also had a deck PDP8E that my partner Rich Alderson had taken apart and couldn't figure out how to get back together. So, I did that, and here you see one of our other colleagues, Keith Perez, who was a hardware specialist and did some of our specialized hardware restorations, including building a disc emulator for some of the older wash tub drives. And so, here's what it looked like. Believe it or not, I gave tours of this, primarily University of Washington students. There's my 780-5 up and running, running the VMS operating system. And it was very popular. And it gave Paul a little more nudge. Maybe it's time to throw a little money into this. Maybe it's time to make this place a little prettier. This was literally a nondescript warehouse from the 1930s sitting in the southern end of Seattle, the industrial district. So, we spiffed it up a little bit. Now, one of the things we approached when we first started developing our stories was we knew that we primarily had deck machines. Part of that just came out of the PDP10 origin story. A lot of the machines we had, the deck machines we had, had come along with the PDP10. It's like, you want that? You got to take that and that and that too. Anybody here who's a collector has heard this story. Deck machines are relatively available because they really did sell a lot of them, something like half a million PDP11s that were, on average, $50,000 apiece. Also, availability of documentation. We were fortunate enough that through most of the acquisitions at the end of Deck's life, we were able to get the documentation from digital, which was a lot easier than with some other companies. Data general almost completely disappeared from the face of the planet. Man, Bruce Ray down in Colorado owed a debt of gratitude for being the person who preserved that and finally brought it public. We also biased ourselves towards interactive systems because one of the things in Paul's mind is how fun is a batch system? Ultimately, we did have one or two batch systems so people could experience what it was like to use punch cards, run the punch cards through the computer, watch the printer spit out paper. As a contrast to the idea that you could simply sit down at a computer and start typing. Paul had a certain set of his priorities. The machines needed to be in operating condition so anything that we couldn't restore, we didn't want. Nothing was going to be just an ornament. They must be available for use by visitors. Again, this goes back to why interactivity was important. If somebody can't just sit down at a teletype and start typing, by the way, the look on somebody's face the first time they press a teletype key and it's completely different than a typewriter or a computer keyboard, it's fun. There was still a focus on machines that Paul himself used. Anything close. There were times that I really tried to get other machines into the story and Paul just said, no, don't care about those. And he wouldn't listen to an argument because it wasn't important to him. I won a few times. He didn't really care about PDP-8s, but they were so important to the story that we convinced him, especially since we already had a couple dozen around. So, again, when we started writing stories, what stories must be told? What are going to resonate with people? How do we connect those stories together? And this gets into a very interesting question about how do you describe progress in computing? There are people that want to talk about it in terms of, well, they got faster or they got cheaper or they got smaller, but those are interesting marketing terms. What we realized we wanted to talk about was how did they get more important to you? How is it that everybody has one in their pocket now? Whereas 30 years ago, probably one in 10 people had ever physically touched a computer. So how do you lead people into those stories, too? And that was an interesting, you know, kind of from both ends of things. We started with antiquity, then we also introduced modern machines so we could connect them back to the vintage systems. And then, ultimately, we're constrained by what artifacts do we have? And those are the nouns in our stories. So, like I said, we cleaned the place up a little bit. This is what the PDP-7 looked like when we got it all up and running. We actually had one of the original experiments on nuclear decay right over here. So that video, that screen, would be able to display the points of data over time accumulated from a nuclear decay cycle. And we had itty-bitty samples that were actually very safe. You could eat them. But still, we kept them locked away. That was one thing we didn't let visitors play with, was nuclear materials. But the machine ran beautifully. I maintained this machine when it didn't run beautifully. Which was very fun. It's all discrete transistors, which in some ways is easier to restore than IC-based machines. Because a lot of the integrated circuits used in the 70s and 80s are no longer made. And I know both the museum and I buy them whenever we can find them on eBay. Oh, a thousand chips for $100? Yes, please. Okay. So I mentioned we wanted to also start from the beginning or the end and work back towards the beginning. So there's the obligatory Microsoft display since Paul was a co-founder of Microsoft. I'm somewhat ashamed to say I put this together. But it was popular. And here was one of our other pride and joy machines. This is Xerox Alto Workstation. This was pretty much the first single-user workstation that was built. There were experiments prior to this, especially at SRI by Doug Engelbart. This took a lot of Doug's ideas, put them into a single box. They made about 2,000 of these. They were never sold. They were an experiment not only in computer science, but also in the social science of computers. Understanding how people would use a computer, why people would use a computer. This is another machine I maintained. Asked me some time over a beer how well I got to know my logic analyzer. So here's the grand opening. Bill and Paul, the gentleman on the left was the business manager of the museum at the time. The machine behind them actually came from Germany. It came from Kiel. It was part of a very elaborate PDP-10 system that got split into two. This machine is why I learned to drive a forklift. When it came in, it was in crates, and the crates were stacked on top of other crates, and there was no way that we could actually get them off each other without a forklift. The machine when we got it was rusty up to about two feet and been standing in some water for a while. You can see that it's actually all nice and clean now, and it works. And there's the group. There were some very illustrious people here, and they let me in too. So one thing that happened during this time period is I decided to take a PhD. And there had been a fair amount of cross-pollination in my own person about what I was doing at work and what I wanted to do at school. So what we were doing at the museum, we had to build exhibits. We had to have something to fill that space. We had to use what we had. We had to figure out how to tie it into other things we had. It'd be very pragmatic about this. At the same time, we didn't want to build just another stuff-in-glass cases museum. I'm not going to name names, but Mountain View has one of those. And personally, I think that's a shame because these machines speak in their own voices. This is the unique aspect of museology is that you have artifacts that have voices and you want to let the machines express them. And talking with diverse groups was so much fun. I put together lectures for fifth graders. I put together lectures for groups of adults. I would have people come through and I would maybe be somebody who was giving them a tour and here I have some of these eight-year-old and here I have somebody who worked in the industry 40 years ago. And speaking to both of them was really a challenge and it taught me a lot about how to connect people with technology even though our society, at least our press, tends to try to distance us. So I started thinking about what does any of this mean in terms of trends and theories and so forth? I was exposed to some ideas such as the social construction of technology that I'll talk a little bit more about as well as value-sensitive design, which I'll talk a lot more about because that is the core of how I'm approaching, looking back and sort of trying to unravel what people were thinking both when they created these machines and when they took to them or didn't take to them. Again, negative adoption is a very interesting coefficient to have. Why didn't people like that? So I had originally started with the idea I'll just do this nice descriptive dissertation about some particular piece of technology and I was going to frame it in terms of value-sensitive design and some of my committees said, oh, you're going to do a work of history and my advisors are like, well, you could be a little more theoretical here, perhaps. Well, I ended up getting a lot more theoretical. Actually, my dissertation is exploratory and applies theory to what we know of history to try to help us know more about history. So I mentioned, I started learning about social construction of technology. This is really key to how I am approaching everything because, again, we try to say technology's over there and maybe we fear it and maybe we think we can control it. Maybe we think we can just make it on demand which is what a lot of our software companies want us to do. Social construction of technology suggests that technology is actually part of us and we are part of it. It is an indivisible construction of who we are. Human beings just use technology and in order to use technology, we have to make technology because by definition, technology doesn't grow in the ground and it doesn't come off a tree. Until we make it, it doesn't exist. So here are some references that I found very useful. So Biker wrote a book called Bicycles, Bake Lights and Bulbs. And I read on social construction of technology. It's very approachable. It has some fascinating stories just leading into the bicycles, looking at the evolution of the bicycle over time and how it changed things like women's clothing because women couldn't ride a bicycle in some of those big elaborate flowing gowns and so women's clothing changed and as women's clothing changed, their perceptions of themselves changed. How they felt in society, how they felt that they could accomplish. Biker Hughes Pinch, usually just called Biker and Pinch, Social Construction of Technological Systems, is a little more formal. And Susan Douglas wrote a fascinating book on radio called Adventing American Broadcasting. And in one of our latest chapters, she says in so many words, in this sentence, this is the social construction of radio. And with all this other stuff I'd learned, I went, I get it. Because radio was really an interaction between society and technology in a way that you can't separate because it would not have advanced but for societal changes and it fomented societal changes that advanced it. So thinking more broadly about history of technology and understanding what it had done previously, an excellent book by Statenmeyer, Technology Storytellers. This is the history, our study of the first 20 years of the journal Technology and Culture. TNC is considered to be the leading journal on history of technology. And this was something that really jumped out at me. He had about five categories of literature that he found. These are the three that are relevant to what I want to talk about. So we have internalist literature. It's about the construction of this thing. It tells us how it was built, who built it, what company, how much it cost, how fast it was. It tells us about the details of the engineering of the artifact. We also have externalist literature that talks about social context, but tends to be very broad in terms of what technology it's talking about. It doesn't include the technology and the story. These are the stories that say things like, the computer influenced the office place like this. The computer. Okay, which computer are we talking about? Which aspects of the computer are we talking about? So there's not a lot of technological detail to tie to the social construction. Then there are books that are contextual. They synthesize technical design in historical context. One of my favorites is by Joseph November. It's called Biomedical Computing. Now interestingly, the ones that I enjoy most in terms of how well they're built tend to be biographically constructed. So Biomedical Computing focused on the life of Wes Clark, who introduced one of the first mini computers, the Link 8. The Wonder Machine was by Waldrop and talked about the life of JCR Licklider. And Licklider was very important to the development of the ARPANET and the internet as we know it. So there was a lot of material to start with, but some of it was kind of disjoint. And it wasn't satisfying to me. It wasn't telling the stories I thought we needed to tell. Well, I did a lecture on material culture. I went to a lecture series. And this is a completely different take on history. Rather than talking to kings and generals, this is about figuring out how to talk to the common man. And the common man was often never interviewed. And so material culture seeks to find about the common man through the common man's common artifacts. So great study on pottery shards and how they tell us the state of a particular culture's technology. Eating utensils, chairs. You know the chair has not always been a common piece of furniture? It used to be very rare and was a sign of nobility. Now, who thinks about a chair? Tell us something about the development of society. This ties in to some of the other things I'm going to talk about in sort of an indirect way because I'm going to talk about reading these things, reading these documents of history that are the artifacts. And Buckland's paper, What is a Document? is actually very powerful in relating the idea of document to more generally anything that holds information. And there's a somewhat infamous paper by Suzanne Brie that asks, Is an antelope a document? And she concludes yes. Finally, Value Sensitive Design seemed like a really great tool to apply. I'll go through this step by step but I want to talk a little bit about what it is and where it came from. So it was developed in large part by my advisor, Batja Friedman at the University of Washington. It was created to promote the socially responsible design of technology. It was created so that you can make new things. What I'm doing, applying it historically is actually pretty new. It inherits heavily from participatory design which Europeans will be very familiar with. Started in Sweden and was primarily about labor relations. Value Sensitive Design sought to sort of expand that to other parts of a technology population. Derived some design tensions, scenario based design and the idea of value dams and flows. There are some values that people must have in an artifact and some values that people really don't want to see in an artifact and so they will promote or completely kill the development. So for instance, just this week we see the Mattel, Politoi, it was an AI companion for your children over concerns that it invaded privacy and it violated the parent-child relationship. So social values caused technological artifact to be pulled from sale. Value Sensitive Design contains three fundamental theoretical components. For one there are stakeholders. Stakeholders are roles in interaction with the technology. There are direct stakeholders who use the technology, indirect stakeholders who don't use it but are affected by it. I came up with a new example other than what I've been using. So for my day job I build spaceships. And if you get in the spaceship and you fly, you are a direct stakeholder. If you're somebody who launches the spaceship, you're a direct stakeholder. If the spaceship falls out of the sky onto your house, it affects you. But you are an indirect stakeholder because you never intended to interact with the technology. It just kind of decided to interact with you. So stakeholders have values or they are affected by values. And I'll talk more about values. It's actually a rather subtle concept. And then there are value tensions which exist between two separate values, for instance privacy and safety. So I wanted to come up with a little one-sentence definition of a stakeholder. A group of entities connected with contextually meaningful similarities. Better is to talk about examples. It could be your neighborhood. It could be, you know, the people in El Paso that we don't want to drop a rocket on. More generally the user or consumer of a technology. What is a value? Last year I attended a workshop in Leiden where one of our key questions and our key research points we came out of a week-long workshop with was how do we really describe a value? So what I have here is it's, you know, something that you can point to. Here's this one value. It's contextually relevant to the technology. So again, privacy and your cell phone. And it's held by a stakeholder. So there is a stakeholder role, a group of people to whom this value is important one way or another. Now there are people that don't like the way we think about values. So the Donatech et al. wrote a paper where they criticized us for being overly prescriptive and actually our opinion is that Donatech over read this. Winner said, oh, value? That's like what I feel like that day. And he was very dismissive of the idea. And we recognize that clearly identifying this is a human value that we can consider in this context is an open question in some ways. Hasn't stopped value-sensitive design for being very effective. So again, value tension. Hey, I have two values. And they interact in a way that's maybe not compatible. So you can deal with the value tension in a number of ways. You can agree to disagree. And maybe what you now do is you've just split a stakeholder role into two separate roles. You can compromise. You can find the sweet spot in between the two values. You can remove both values from discussion, like Mattel did. They simply said, we're not going to make this. This value tension cannot be resolved. So what do you do with all this? Now we've got stakeholders, and we've got values, and we've got value tensions. Well, there's a three-part methodology to value-sensitive design. It involves conceptual investigation, empirical investigation, and a technical investigation. These three aren't in a particular order or anything like that. This is designed so that you can iterate through them. You can flow through a given investigation as many times as you need to. You can start with an empirical investigation. You can start with a technical investigation, which is what I'm doing. So the conceptual investigation is where we talk about stakeholders and values. And where we get that information is through an empirical investigation in which we connect these ideas of stakeholders with persons, and then we find out things about those persons. And in the case study I'll discuss, that will become very clear. Technical investigation, if you're making something new, is now I know all this stuff about the stakeholders and their values, and I've talked with them, and now I can take what I know about that and figure out how I can change technology. So I can't make it float in the air with anti-gravity, but I can do a lot of other things. So what technological aspects and affordances and features will allow me to express these values in this object of technology, in this artifact. So I stole this from another presentation, so pardon me that it looks really weird. Value-sensitive design essentially says this. We start with an idea. Somehow we're going to get over to techniques and things. And as Layton suggests, we do that through the arc of design. So what I asked is, can I take that idea and turn it around? Can I say, I'm going to use this as an historical lens to understand how we got here from there? And I'm focusing on conversational interaction with computers. I wonder where I got that. It was probably living computer museum. So I'm effectively doing this. Starting with my things. Figuring out the design elements from the evidence of history and using that to get back to the original ideas. So again, this is how I'm approaching value-sensitive design. The fundamental idea that I propose is we look at the artifact because we have it. Or we have had it. Okay, what do we know about its features and affordances? So now I do an empirical investigation to go find out who used it. And then, from that information, I do an analysis to find out, all right, how do these people group out into roles? And this has been really fascinating in the case studies I've done, understanding how even we define the roles, we find the roles, we do not define them. We find them. And the values of the people in those stakeholder roles. So I'm doing 3K studies in my dissertation. And one of them is on the Dartmouth time-sharing system. This is the one that's most mature. It's still a work in progress. But I think this is probably the most helpful in terms of really talking about these sort of abstract ideas. So technically, the DTSS system, the Dartmouth time-sharing system, was effectively an asynchronous multiprocessor. There was a GE 235 mainframe, very conventional computer at the time. And then there was this device called a Datanet 30 that was supposed to allow you to submit your batch jobs over telephone lines. So this was non-conversational interaction. Yeah, you could send a job down. You could find out when it's done. But it was very limited communication. And the point was that you didn't have to have somebody running around plugging things into a switchboard to deal with the 40 terminals on campus. So they wanted to build a time-sharing system around this. They being doctors John Kemeny and Tom Kurtz at Dartmouth University. So they wrote some software. They programmed the Datanet 30 as a master computer that pushed little bits of work down to the GE 235 and would then route them back out the Datanet 30 to the person who had asked in a way that you didn't know, that there were 39 other people doing exactly the same thing. Now, this would have been very cool, but there were some time-sharing systems available, or in the world. These were still not commercially available machines. But they wanted to make it so that everybody could use it, and not everybody wanted to learn machine language. So Kemeny and Kurtz created a language called the Beginner's All Purpose Symbolic Instruction Code. You didn't know that had a meaning, did you? The basic programming language, the first instance of it was on the Dartmouth time-sharing system. And they eventually added Algol, Lisbon, Fortran, as well as an assembler language. But basic was by far the most popular language on the system. So I mentioned there were other time-sharing systems. MIT had the compatible time-sharing system, and MIT and Bell Labs evolved that into something called Multics. Interesting thing technologically that's different is that both of those machines worked by interrupting the threat of control of a program. So I'm going to work on this program, and now I'm going to work on this program, and now I'm going to work on this program, whereas the DTSS system took anybody's job and just split it up and multiplexed it into the system. The system itself had no idea that it was doing time-sharing. So in the empirical investigation, I'm using a combination of literature and interview. The inventor designers, this is a very simple population. John Kemeny, who is now deceased, and Thomas Kurtz, who is not, fortunately. Then there are the users. Now I should clarify, inventor designers are not stakeholders because they are the originators of that artifact. They add their own values, but we distinguish them from stakeholders because stakeholder roles can merge and change and bifurcate. Now the users for this system were exclusively university students, local and remote, and high school students. The machine came online in 1964. By the end of 1964, there were half a dozen high schools connected to it with teletypes. So two parts to this empirical investigation much of it I have to do textually because some of these people are no longer with us. One of the really wonderful resources for this has been Kemeny's book, Man and the Computer. He doesn't really describe the Dartmouth time-sharing system. He describes his dreams for the Dartmouth time-sharing system. He talks about how he sees the role of the computer in society and you can read through that how he came to envision and build DTSS. There's also still a lot of technical documentation available including from GE Corporation itself, the people who were providing the hardware. There were reports back to his funding agency that were very detailed, scholarly publications and then some books that were written after the fact and Bulls is very, very highly respected. It mostly deals with the immediate successor to the original system or what they did is they replaced the 235 with the 635 and they went from supporting 40 people simultaneously to 200. So this is the bulk of the literature I'm reviewing. Then I'm also using semi-structured interview. I love semi-structured interview because it gives you guidance. It gives you a shape to the interview so you don't just end up running on. However, it lets people run on. And some of the best interviews I've had with people are where they went off the text and they started talking about what mattered to them which is really what I wanted to know. So I had the opportunity to exchange email with Thomas Kurtz. He and I have had three or four email exchanges and I was very surprised to get his email in my inbox. Then the other group has been high school students. I have been unsuccessful at finding anybody else. The system maintainers and operators for DTSS were students themselves. It was a school project. It was not part of the university per se. I have had a very difficult time, an impossible time, contacting university students. Dartmouth had a celebration for DTSS a few years ago and now seems to have completely forgotten about it. I can't get them to return phone calls. So I also can't find out if there were any faculty adopters. I can tell you that at the high school where I found people there were no faculty adopters. The kids were on their own which was a fascinating story. So I'm learning about Kemeny through literature, Kurtz through email discussion and GE itself is a very important inventor designer here because they built the underlying platform that enabled Kemeny and Kurtz to do what they did. I'm considering them part of the investigation. So what kind of values did I see? Well, Kemeny and Kurtz were very strong on accessibility to students. You notice I've got this broken out. I'll talk about that in a moment. Underlying all of this was inclusion. It was important not just to have a system that made the computer science department run better. There was no computer science department. Kemeny and Kurtz wanted a system that could be used by math majors and chem majors and physics majors and liberal arts majors and whoever else wanted and believe me they did. The historical record shows that there were many, many programs written that had absolutely nothing to do with computer science but you've made use of the power of the computer. So here are my stakeholders. The student users, the university students I can't find, the high school students I can't find, and I've interviewed about nine of them. They were all from the same school, Philips Exeter. I invited all of the schools who were listed in the documentation and Philips Exeter is the only one that's really were applied. Nothing from faculty and again I can't find these system operators or maintainers. Now here's something that's very interesting about understanding stakeholder roles. Sometimes you think, well this is who they are, they're these students. Now actually there were two distinct groups of these high school students. There were technologists. I mean these were the people that would boast to me about how they hacked the system. They would boast to me about how they learned so much about everything that was behind the curtain with DTSS. And then there were the application developers. They used the system for math or science or electronics or whatever. Some of them actually shared some of the code with me that they have saved over 50 years. So it's really important to make sure that when you're looking at your stakeholder roles, you've really thought through presentation one of my colleagues did. He said, okay law enforcement. And we ended up with five stakeholder roles that composed law enforcement. Now I'm going to go back to the empirical here for a moment and talk about the interview population. Phillips Exeter was a boys boarding school. It was a very elite school. The school was completely focused around the idea of individual curiosity and discovery. I learned this through interviews. I also learned through interviews that these kids that are now my age and older still have an incredible loyalty to their school. This shaped their lives. And they tell different stories about how it shaped their lives. But none of them came away from Phillips Exeter unchanged by their own words. So now let's go back to conceptual here and let's talk about values. This is preliminary. I have not done a detailed deep dive yet. But in the technologists, they found this sense of wonder. Look, this is a computer. Look what it can do. And they were digging through to find everything they could make that computer do. And there was just simple curiosity. Boy, what could I do with this? And it was more of an abstract what could I do with this? The application developers were much more concrete. And by application development, they're writing 100, 200 line programs but maybe to do something like analyzing electronic circuits or something like this. But they felt very empowered to ask questions they would never have asked before. This came up from several interviews where people said, you know, if I had to sit down with pencil and paper to do that, I would never wrote it on that. They had this computer. The other value that came through was intimacy. And you're saying, OK, what does intimacy mean with a computer, with a teletype? They felt a connectedness to the conversation, to the interaction that they were having with this machine. Some of them would spend literally hours in this little closet with a teletype. Now, keep in mind, they did not have any coursework that made use of the computer. There was no curriculum around this. They taught themselves and some of them would literally have to be kicked out of the building and told to go to bed. The other thing that came through was an individual commitment. So the machine felt, they knew that this machine was hundreds of miles away and was being used by a lot of different people, but no, it's mine. I have it all to myself. I have this multi-million dollar computer all to myself. There was this sense that it's mine. We're on the same page. So my preliminary summary is the system did its job in terms of supporting computers in education as Kamini and Kurtz wanted. It was also valuable to GE. That's one reason that they were involved with the hardware, is that this is how they designed their own eventual commercial time-sharing system based on the research that was done at Dartmouth. And our key stakeholders found that the system enabled their values. The technology guys went, here's a fun new toy for me to play with intellectually. The application developer said, this enables me to think in ways I've never thought before. So I need to find representatives of these other roles if I can. And I might not be able to, and I'll talk about that in a moment. I have a lot more work to do on the interview corpus. I have a lot of interviews. Sometimes it was an hour and a half, and I'd have to tell someone, I'm sorry, I need to go. I'd love to have more of this conversation. And then I'd get five emails from them within the next three days. People loved telling this story. I asked, but couldn't find out much about non-adoption. So here you are on campus. Are there other kids going, why are you spending all your time in that little room? Who needs that kind of stuff? No, most of the kids were purely oblivious, which is another category of non-adoption. These didn't know. They had their other curiosities. They didn't need a computer. So to wrap up, I want to contrast the other two case studies I'm doing. I'm looking at Multix. And I actually looked at Multix versus a counter example because anecdotally, I thought of it as a programmer's system, so it was going to have a very narrow focus that was not going to really connect anybody with the machine conversationally. And so far I've been wrong. That's what research does to you. And for instance, there is the Multicians.org web page. And there are pages of people who use this system, many of whom you can still contact. There is a lot of community around the history of the Multix system, even though the last one was decommissioned in 1999. The other one I'm looking at is Xerox Alto. And I found that interesting because it's still a conversational interaction, but in a non-textual sense. So much of what you do is moving little pictures around the screen. And it's a very different form of communication, but one that can be even more intimate, one that can draw one in even more. And I found a very diverse stakeholder set there. And I have some hypotheses on that. But it's something that I definitely need to explore more. And I did not expect this. It's a little surprising. And I had even suggested maybe I should focus on just one case study. And my committee came back and said, no, you've got a great question by the tail. So that's going to take some more delving once I have some more interview content. But that's the question there is this system had these stakeholder roles, this one these, this one these. Some just don't exist. Why? It's a lot of fun. So thank you. Glad to answer any questions. I guess that's going to be fine as well. Oh, there we are. Oh, now it's working. Now it's very loud. I was just wondering because, because I'm not the super duper insider. It's super interesting to follow your questions, but especially regarding design, you wouldn't have images of the interface or of what it means to work with these kind of networking computers. Would you? Because it's very, you know, my imagination then is a little bit lost in all these interesting thoughts and words because I have no idea how this would look like. If I understand your question correctly, you'd like to understand how people were all connecting through the computers? Yeah, to see the, to get a visual idea about it, you know. Yeah, see the types, see the interface, see the programming language. I don't know. That's a great question. I wish I had something with me, but I didn't think I was going to have three hours. Just very briefly though, when you look at the time sharing era, which was about a couple of decades, it was really about a machine that has very little intelligence in it, whether it's a video terminal or a teletype or something, that's connected to a central computer. And so it's a star network. Everybody's connected to this one computer and everything goes through that computer. And you can send email by putting it over here on the main computer and saying, I want that person to have it and there it goes. You can have chat conversations. You can do collaborative work. There was a system called ITS, the incompatible time sharing system created by MIT, that was really the first social workspace. It had no passwords. Everybody could see everybody else's work, and so it allowed them to collaborate freely because that's what they believed. That was their fundamental value, was that data should be shared, not hidden. And then when we started to break systems up into workstations like the Alto, one person, one Alto, we had to have some way to reproduce that connectivity and that was why the inventors of the Alto also built the first Ethernet. That was where Ethernet was invented. And now that becomes a bus network that's become expanded phenomenally, where everybody's talking to this common carrier and a signal goes out and goes to everybody and the right person grabs it and does something with it. So really that's kind of the fundamentals of the two structures and there were some hybrids that came in the middle using various different schemes of serial connections and so forth. Those are the really fundamental drivers of networked conversation between computers. I hope that helps. Thank you. I like your approach of interviewing all the stakeholders and especially the users of the computers but as a person who works in a museum I also like your idea of every artifact has a voice. So in your dissertation do you also have research on that voice of the artifact itself? The shape of the design of the artifact itself. Do you also have a research on that or do you rely on the interviews? I do have some theoretical research on that where I explore more of this idea of intimacy as a value in human-computer interaction and I think that may get somewhere near where you're talking about. Not certain but that's a great question. I'm going to have to ponder that a little bit and by the way one thing I really don't want to forget to do is to thank Eva for inviting me here. It's just wonderful to be here in Berlin. Anything else anybody would like to bring up? In terms of user interaction with these old computers what strikes me so fascinating about older PDPs is the really, really awesome product design. The color is just perfect for the era. It's super fashionable in terms of shades so I'm wondering how much accessibility via the product design of a PDP was there in the late 60s, early 70s. I mean I look at this computer and think, oh awesome I totally want to have this in my office. Would normal people actually have seen this back in the day and recognized it as some awesome design or the color scheme and things like this? It's a great question. I know that there were some businesses that were very forward about their computer rooms. They were not back room things and I think definitely played into that with their color schemes and their aesthetics. There is a fair amount of work about the aesthetics of technology design. It goes beyond functionality. It goes to how do you appeal to perhaps a non-domain expert? By the way I'm glad you used the word accessibility because I thought I had a slide on this and I guess I must have deleted it. I used a couple of terms earlier that I should explain. DTSS supported accessibility in a couple of ways. I break accessibility into three separate values. There is structural accessibility, which is can I get one? So when computers are $3 million, very few people are going to have any kind of access to them at all. And one of the things that Kemeny and Kurtz wanted to break down using time-sharing was the per-unit cost, if you will, of that access. They also focused on conceptual accessibility and that is once I have my hands on one, can I do something with it? Is it something that I have such a steep learning curve on that I'm just going to walk away? I don't care. The basic programming language was developed to be very English-like, to flow very easily, and to be pretty foolproof. They put in a lot of great error messages. Hello software industry, you can do better at that today. And that was a big factor in its adoption. The fact that there was no class for it. There was a 25-page manual that outlined the basic. And that was how these people learned how to use it. One of them went on to work for 34 years with Intel. Then what we commonly talk about in terms of accessibility is what I call physical accessibility. It is, is there something about this system that impedes my use because of my physical limitations if I have some kind of a disability? And there are so many disabilities, not just the ones we see. You know, here's mine. When I was at Microsoft, we developed a phone, a pocket PC phone. And it had this huge screen. It's as big as my Samsung S6. And when the phone would ring, the number would show up in these tiny little letters. And at the time, I didn't wear bifocals. I just wore reading glasses. And I had to pull my reading glasses out to tell who was calling me. So that is a flaw in physical accessibility. But I wanted to call that out because I think that that is a really useful distinction. And systems like DTSS did a really great job. But it wasn't until, fair amount later, when adoption was broad, when structural and conceptual accessibility were very solid, that we started to encounter a population of people who didn't have the same physical abilities as the multitude. But yes, the deck designs, it's interesting if you look at some of the literature of the time, the advertising literature, very much was focused on not just the businessman, but also just catching people's eye. One of my favorites is a PDP-8 mini computer sitting on the ground. And here is a woman's leg in a go-go boot, just top of the computer. It was hot. Anything else? Basic history. Thank you. Where in the time frame with Gates and Alan and Washington State, were they pre or post DTSS? They were post DTSS. I don't know, and it would be fascinating to find out whether either of them really had encountered news of that technology, because basic really became very popular very fast. It disseminated very, very quickly. There were tons of variants of it. And the PDP-10 basic that they worked with was certainly a variant. So I think that was probably their first interaction with it. But they had first been on Genie, which would have been the GE system, right? But that was a time-sharing system, and it was only in use for about a year. Okay. And so that was not the DTSS piece that came? No. That was probably, I think that would have been running GCauss. Great. Which was GE's commercial version of DTSS. And I should mention that there is a DTSS emulator available online. This is how much passion there still is for this machine. And it's really fun to run the DTSS basic, because it's very simple. It doesn't have all of these bells and whistles that we bolted on the modern basics. It's just a really straightforward language that actually lets you do a fair amount. It has floating point and it deals with strings pretty well. It has good control structures. Yes, it's not surprising to me that people say, wow, I want to play with that. Thanks for the question. Is everybody else thinking, gee, I want a beer? Just a moment. I got another question also picking up on your last slide about why are there different stakeholders? And of course you take a technology-centric approach. But also putting it in context with the question on product design, I would suspect that just for instance taking the Alto, and everyone knows of course the Alto was kind of a blueprint for later Apple computers. Of course these computers were marketed very differently. They were designed differently. The organization behind it was different. And I would suspect that all these factors play a role in how they shape stakeholders as well. Is that correct or am I just short-sighted here? I completely agree with you and that is actually why my research stops before the broad introduction of what I call the consumer computer. So there are people who want to talk about the personal computer and just reading something earlier in the program for this event that talked about the LGB 30 as perhaps being the first personal computer. There are enough definitions for personal computer that everybody can have their own. So I like to talk about the consumer computer, which was something that was mass marketed to people who were not necessarily going to do anything important with it. It was just, oh, you need one of these. It was the kind of beginning of that marketing. And of course that required aesthetics as well as functionality. Deck, I think, was probably on the forefront of that. Data General as well. They made some very, very fancy designs. Some great color schemes. So, yeah, that's an interesting factor is when you throw in that marketing angle and how that shapes stakeholder roles. Briefly and very roughly, because I'm just starting to think about this, the hypothesis that I have about the earlier era is that there was greater knowledge, there was greater exposure to the idea of the computer as we came from the days of DTSS to the days of Alto. The people I've interviewed who worked with the Alto, very rarely do you find that this was their first computer experience. Now, high school students that worked with DTSS, almost exclusively, not exclusively, there was one who actually worked with IBM 650 in his first year of high school, which was not at Phillips Exeter. But I think that there was a difference in the exposure of people to computers. Both through, and this is another part of my dissertation, I don't have time to talk about it unfortunately, both through the press, factual literature, and I think also through fictional literature, science fiction. Science fiction can be a very powerful tool to elucidate the values of an era because the authors of science fiction and the inventor designers and adopters of technology all lived in that same social construction. And the science fiction authors did something else with it. They just ran with it. And in fact, there is a formal tool in value-sensitive design called the value scenario, which is, in essence, a structured fiction about what your new technology will do. Sounds a little bit like science fiction, doesn't it? That paper's in press. Anything else I can answer? Well, thank you so much. My applause for the audience. I'm grateful to all of you for being here after this long and full day. Thanks.