Molly Wright Steenson: "Architectural Intelligence: How Designers [...]" | Talks at Google

Molly Wright Steenson:

Show Video

So. Hello to friends, in the room friends, who, I know are joining us virtually I'm really, excited that you are here and thank. You Justin for having me it's great to be here. This. Is part of a big, bunch of research that I started, some way or another 13. Years ago and some way or another in, 1996. Even, and I. Want to start by making the claim that architecture. AI and design are all connected. These. 60 these connections. Go back some 60 or 70 years to 1956. Or even 1948. And. Newness. Seems to be central, to how we talk about artificial, intelligence, today. We. Say things like AI is the new black. AI. Is the new UI. Artificial. Intelligence, is the next digital frontier, according, to our friends at McKinsey. Andrew. Ings stated, that AI is the new electricity. But. Also AI is the new MOOC, because. He will be an online course that he's running. You. Need data for AI so, data. Is the new oil. The. Future of computers, is the mind of a disembodied toddler, head. And. Apparently. I am also the. New AI and I, will spare, you this, what. This says but it's really really sexist, and really awful if you. Want to get an idea of how much the word new gets used when we talk about AI you could look at something like this MIT Technology, Review article. I just, highlighted. The word new where. I could find it and there. Were six mentions, and the first two paragraphs of course I think one of them was New York so I don't think that that actually counts. And. This. Gentleman was. A part of a conversation about tech muggles declaring, an era of artificial. Intelligence. But. AI isn't the new because AI. Isn't new, here's. How new it's not 1955. John. McCarthy the. Founder, of the. Stanford AI lab and before that the MIT AI lab. Referred. To Artefill coined. The term artificial intelligence, to mean making machines do things that would require intelligence. Have done by man. In. 1956. He brought together a group of researchers at, Dartmouth College in New, Hampshire and they, spent the summer hashing, out what they thought the platform, of research on AI should be and it, included, things like neural. Nets and it included, what. Did what does he say says an attempt will be made to, find. How to make machines use language form abstractions, and concepts, selves. Kinds of problems now reserved for humans and improve themselves we. Think that a significant. Advance can be made in one or more of these problems if a carefully, selected group of scientists, can work together on it for a summer. 1956. Herb. Simon Alan Newell and JC shot, Simon. A Newell very important to Carnegie, Mellon's history, wrote. In 1958. That they'd pretty much figured, out this problem of modeling the, human brain in, computing. They said intuition, insight and learning, are no longer exclusive, possessions, of humans any, large scale high-speed. Computer, can, be made to prote and made to exhibit them also they believed that by the 1960s, they'd, haven't figured at a figured, out and. Of. Course Marvin Minsky who in 1961. Said I believe that we are on the threshold of an era that will be strongly influenced, and quite possibly dominated. By intelligent, problem-solving, machines. Proto. Machine, learning, you. Might find in, the, world of perceptrons, in 1958. The. New York Times published. An article talking about this new Navy device that learns by doing and said. That the Navy revealed the embryo of an electronic, computer today that it expects will be able to walk talk, see, write reproduce, itself and be conscious of its existence, the Navy, said the perceptron, would be the first nonliving mechanism, capable, of receiving recognizing. And identifying, its surroundings, without any human kiman, training or control. So. All of these ideas are, really old. They're. Older than me by far and they're. Older than a lot of what we work on but. In it to this I want to point out that AI and architecture, are old friends and kind, of co-created, each other at a certain point in time that in ways that has very. Direct, impacts.

On The kind of engineering, and design work that we do today and the problems that we focus on, I'm. Going to introduce three of the characters in my book. Art. Of architectural. Intelligence, the three people I'm going to introduce in, this talk are Christopher, Alexander, Cedric. Price and Negroponte II Nicholas Negroponte II who here is has heard of Christopher. Alexander. Couple. Anyones. Cedric price, right. On and, Nicholas. Negroponte a. Couple. More all right so, they're. All architects, worked very closely with. Technology. In different kinds, of ways some, so much so that you don't probably think of Nicholas Negroponte II the, founder of the MIT Media Lab as an architect, and, I'm going to talk about the ways that they worked with cybernetics. And artificial intelligence to, build new kinds of worlds and new kinds of approaches, in ways that affect what we do today, I'll. Start with Christopher Alexander. Christopher. Alexander, if you do know of his work and even if you don't buy me putting up this book right here I've probably, begun to bring him into something you've heard of before he's, written many many books he's, mathematician. Born in Vienna raised in England came. To the United States in the 1960s, he's, quite old but still alive living in England again was a professor at Berkeley and, he. I'll. Mention three books today notes, on the synthesis, of form a pattern. Language in the timeless way of building notes on the synthesis, of form was his his, thesis, and if you've ever heard the idea of design being. Good. Fit or fitness. You. Know trying to shim. A table, until it kind of works or putting, a stable, system in place that's where that idea comes from those. Of you who have used patterns, whether in UX, or in, software, are, are. Right in the lineage of Christopher, Alexander, his idea of pattern languages, have influenced. This and then his philosophy, of design which he outlined in the timeless wave building if you've, ever used a wiki or you use agile processes. You. Have used something that that. He's done my. Own research with, Alexander. Started when I was reading notes on the synthesis, of form and. This question of a I came up when I realized in a footnote he, was referring to Shannon. And, Claude. Shannon John, McCarthy, and Marvin Minsky, and I wondered why is this architect, what. Does he have to say about artificial, intelligence in 1964. Thirteen, years later I have, a book about that. He. Was interested in finding ways to break down design problems so they could be easily understood and, designed. Anybody. Who has created tree diagrams, in order, to explain, the function of a piece. Of software or website, is working, in this lineage. In. Fact this is the outputs, of one of his programs from this is a something. That's in the Berkeley archives that shows how. You'd break down the elements of a design problem for how you design highways. He. Also applied these ideas to the BART system in 1964. This is, 390. Requirements. That he, and his team outlined, and this, is before Bart was built, I'm. Not sure how many of these are in place today I think we probably need new requirements, my ears are still ringing from the BART ride I took this morning. Pattern. Language and timeless way of building I already mentioned, briefly but I want to tell you a little bit more about them. The, pattern language as he puts it and it's not the only it's a pattern, language it's, 253. Patterns from large to small scale and then, the timeless way of buildings is a philosophy, of patterns and these patterns are anything. From how to make a nation-state, down, to how to organize your bedroom and. It's. He in. His and, his, collaborators, write each, pattern describes a problem which occurs over and over again in our environment, and then describes the core of the solution of that problem in such, a way that you can use this solution a million times over without ever doing it the same way twice.

So. In the book here's. An example of a pattern I live in Pittsburgh and, I live in a row house and, so it's, very dense, where I live and these are patterns about the design of a row house you see. Number statement a picture, a. Statement, about the design problem, illustrations. To carry, out to to, explain some of what that design problem is doing and then, a connection, to other patterns, in. The system and it works sort of like an operating, system the patterns are a network, so you see here are a number of them in their, order explaining, what each of them do, you. Know this is going from a mosaic of some cultures all the way down to a household. Mix and it goes further. And further into more detail. Alexander, is someone. That a lot of technologists. Have read if. You know Alan, Cooper the, interaction. Designer he. Used to ferret out Christopher, Alexander's, books in his high school library and, read them there he, wanted to be an architect, so did Kent Beck and, then realized, he didn't want to build buildings, he wanted to build software, and there was a lot more that you could do, within. That world and. Ward. Cunningham, Ward, Cunningham and Kent Beck we're. Curious about how they could apply this idea of patterns to the idea of object-oriented. Programming, languages so they applied it to interface. Patterns, in small talk in the late 80s in about 1987. And a community of programmers got. Together and over the years built up what was called, the. Design patterns. Movement. And this group, of four men who called, themselves the gang of four are, the, ones who created this kind of network of design. Patterns in software, and, it's an idea that you see millions. And million tons of times I think there's something like twelve hundred bucks on Amazon, that refer to design patterns, in in. Software, and in interfaces, and games, whatever. Okay now does this how many people have heard of design patterns there. You go. Also. If you've ever if, you've ever used agile, processes, or, followed. Extreme programming, Kent. Beck and, board. Coming Cunningham, and a number of other people applied these ideas about the, philosophy, of design. From, Christopher, Alexander, in the timeless way of building to, up end. Change the politics of the design process and to, give more flexibility, to users, into people and when I interviewed Kent Beck he told me that it was a rearrangement of the political power in the design and building process.

Everyone. Here has used a wiki and, the. Wiki format was developed by Ward Cunningham, he was developing it in HyperCard, in the early 90s and then. It. Was suggested that he could maybe build this on this new thing called the World Wide Web and what, he wanted to do was have a conversation, that never had an end if he wanted to map all of the knowledge he had of his that, he and his team had he, didn't want to say it ends here he wanted it to be able to go and go and go he, never patented, the. Wiki, format and, it. Was picked up in the early 2000s. To run Wikipedia, and now we've all used wikis one, way or another but, this is again a direct inspiration. Or, a, direct, influenced. By Christopher, Alexander, project. Okay. So that's Christopher, Alexander, the other thing to say about him is architects, tend to hate him and technologists, tend to love him I'm not sure why conundrum. But kind of interesting, so, I'd like to introduce you to a different kind of architect this is Cedric price. I've. Called him the secret patron saint of interaction, designers, he changed, the way, any. Number of people in the UK understood. Architecture, and buildings and what they could do his. Life partner is, was. Eleanor Braun he died in 2003. Eleanor Braun was in the Beatles movies and was. Responsible actually, she was the inspiration for the song Eleanor, Rigby so. Just. A very curious and very funny person, who liked to ask things like technologies. The answer but. What was the question. And. Cedric price was best known for, things. That were never built so, for instance the fun Palace was designed, in the 1960s. And it was supposed to be a big cybernetic. Theater. Movable. Kind of place where all sorts of things could happen and, the. Space, was going to learn from its users over time and adapt to their interests, and means and you could change it over time and use it kind of as a big, leisure. Center to learn and, experience whatever it is you might want to learn and. He worked with the 25-member, 27-member, cybernetic. Community, which is what you see here on the right and in, this particular image it's kind of hard to to read but. It, talks about, input. Of unmodified. People, and then output of modified, people, so they thought through this like a system, diagram, he. Did this project with Joan Littlewood who is a radical theater director, and a protege, of the German playwright Berthold, brushed. He. Also was really inspired by Gordon. Pasque and Gordon Paz is a cyberneticist. Who. Also had. A big influence on a lot of a, lot, of architects, and a.

Piece That he wrote in 1969. Talked about changing, this design. Process. And some really interesting in, some, interesting ways and. He. Suggested, that we turned, the design paradigm, in on itself and let us apply it to the interaction, between the designer and the system she designs rather, than the intersection, between the system and the people who inhabit it, so, rather than the. System you designing doing your bidding maybe it does something different maybe. It meant, s-- and creates, something different. Than, what the two people put together. This. Is a trying, he gave of second-order, cybernetics. He, wrote things like assume, for the moment I'm the successful, businessman, with the bowler hat and I insist that I am the sole reality, where everything else appears only in my imagination I cannot. Deny that in my imagination, there will appear other people scientists, other successful, businessman, etc, as for, instance in this very conference, since. I had the since, I find these apparitions in many ways similar. To myself I have to grant them the privilege that they themselves may, insist that they are the sole reality, and everything else is a condition, of their imagination, on the other hand they cannot deny that their fantasies will be populated, by people and one of them may be I with, Buller hat and everything. This. Is second, order cybernetics. Where the fact that you're engaging with the system changes the system and if, you start looking at the design process is something that needs to take that into consideration. Design begins to look a little bit different. Subject. Price was, again, very, very inspired. By his interactions. And collaborations with, Gordon Paz and one of the buildings, that or one of the projects, that he designed was something called generator, this, was never built but. It was a project that was in existence from 76, to 79, and it was a set of 12 foot cubes 150. Cubes walkways, boardwalks. That, could, be moved around and, and. Recombined. For whatever purposes, someone might like this was an arts retreat, center for a wealthy patron in Florida. Kali's. Really liked moving, cranes so, there's always a moving crane, here and, generator. Would come together like this there was grid on the ground you'd, put cubes over it, roofs, on top connect, the parts and then. You could put up the side baffles and the. Stairways, and do. What you'd like with the space you could also model it with these little qubit to parts. That. You would see he, thought that may be fun things might happen like you could take a bleep or walk I won't mind you this is, 1977. Or so when this idea came up and. Something strange about a mouse rolling around a cube but. It's. Strange. And weird and funny all the time and. Again. We look at Gordon Pasque Gordon. Pasque did. A project first in 1953. And then in 1968. Called which. Became the colloquy of mobiles. And what, happens with these Mobile's are you interact with it and in an inner Anor it interacts, back with you, until. It gets bored and Paul. Pangoro who, is a cyberneticist. And, and, a professor at the College for Creative Studies in, Detroit has just. Rebuilt. The. Kulluk way of mobiles, for the 50th anniversary of the original project he is Gordon tasks archivist, and it's. Up right now as we speak in Detroit. So. This is what. What does this have to do with generator. Well-centered, price realized it was unlikely that people were going to want to move around generator, enough and learn from it and do, the kind of funny things that it would expect so he began, to work with John Frazier and Julia Fraser who were computer, scientists, well as architects, and. Wanted. To put. Together set of programs that people could use that would be used, by generators, so there would be microcontrollers. On all, of its parts an, inventory program, a design program that's what you see here you can pick up and move these cubes, and they plot and printout and. A, boredom program, and the. Boredom program, said. The, following in, the event of the site not being reorganized or, changed for some time the computer starts generating unsolicited. Plans, and improvements, and. This. Is something that John Fraser wrote in his letter to cedric price introducing, these programs if you kick a system, the very least you would expected to do is kick you back and.

Then. He. Said you seem to imply this is a handwritten footnote, that you can kind of see down here in the letter you seem to imply that we were only useful if we produce results that you did not expect I think this was this leads to some definition. Of computer aids in general at least one thing that you would expect from any half-decent program. Is that it should produce at least one plan which you did not expect. So. This is where something like Gordon past's colloquy, of mobiles makes its way into. Cedric. Prices architecture, and these. Are ideas that we pick up today in some, really lovely ways on, the next slide I'm going to introduce a, project, by the, roboticist Madeline Gannon she just finished her PhD at, Carnegie Mellon in. Architecture. And she did an exhibition at, the, London Design Museum with. A robot, a robotic, arm that would. Play. With it's play, with somebody and its midst until it got bored and went away thing, comes, together and, you're in the space with the robot and you just have a very raw experience with, this, animal-like. Machine, responding. To your every move all the technical aspects. Sort of melt away into the background. It's. Incredibly, important, to have opportunities. And spaces, to. Come, in and experiment, and misuse. These. Existing, technologies. She. Didn't know about burton pastor cedric price when. She, did this project but i find it compelling. That many years later this. Is exactly what she is trying to do and trying to get us to rethink our relationship with our machines she. Calls herself the robot whisperer. Third. Person i'd like to talk about is Nicholas Negroponte II and. You'll. Notice here he always seems to have his fingers in the picture, he always seems. To be gesturing, at something. You. Would probably know him as the founder of the MIT Media, Lab but, before, the Media Lab he. Ran something called the architecture, machine group at MIT, and that's, what I want to want, to talk about here, the MIT, architecture, machine group was in the school of architecture but, worked very closely with, the AI lab and developed, interfaces. Programs, and environments, for. For. Artificial intelligence. The. Gesture thing he he's. The author of a number of books the. Book that you'd probably be most likely to know would be called being digital, and being. Digital came out in 1995. But. In 1970. He wrote a book called the architecture, machine and he dedicated it to the first machine that can appreciate the gesture. Nicholas. Negroponte he is a clever man but, this book I think is actually. Fascinating. And really important, it was a theory, of design. For, artificial, intelligence that, I think still holds weight today I. Want. To point out a little bit about the, funding structures, of the architecture, machine group and of. Artificial. Intelligence, in general, the. Architecture, machine group was funded, by the information. Processing, technology, office of DARPA and the Office. Of Naval Research the, office, of naval research was, so important, for defining. The, funding. For artificial intelligence from. Pretty much its inception. Onward. Through the 80s, so. Vital that I, think, that Marvin Minsky, has referred to. Marvin. Denna cough who was the the program officer as the, grand old man of, artificial, intelligence so.

The. Thing to point out is something that caught that on Paul Edwards the historian has referred to as the closed world that artificial. Intelligence. Was. Primarily, developed within, Department. Of Defense funded entities, because. It, meant a small, group of people could continue to work on these projects, through the same social, network friendship. Networks and professional, what networks that had been developed since the end of World War two where. NSF, funding, National, Science Foundation, funding kind of automatically. And invites, very spirit tried to throw things open to the world and open things up and bring more people in, so, it was desired, by, DARPA. DARPA, tiptoe and onr, to. Keep, funding, and development, in these small networks. Of people to. Get things done better and it. Meant that the. Architecture, machine group started, to structure their work like the AI lab did at MIT, and work very closely on. Interfaces. And small. Machines and environments, for artificial intelligence. One. Of these first projects, was funded by IBM, it was called urban 2 or. Urban urban, 2 and then later urban 5 and it was a conversational, user interface, for urban design this. Is 1970. I actually, started, a little bit before that and. The. What happens here is user chooses. A set of blocks you see these squares down at the bottom and then, there's a conversation going on at the side using this light pen to select questions and answers and, it. Asks you questions and, you, you, use the conversational. Dialogue and the buttons here, to. Determine, what. The attributes are, of these, these. Areas, that you are designing, so it's for urban design and. The. Architecture, machine group book, or. The architecture machine book outlines and. Critiques. There there study here. And. You might notice this last one that's. Ted. Many conflicts, are occurring. And. They said you know it. Was charming, but it printed, garbage, but at least it was friendly garbage, so the words that you they used and. It. I think it points out if you anyone here has tried to has. Tried to design a conversational. User interface, you know how easy it's going to seem on the surface and how, hard it is in, reality, and even, just in what we could look back to someone like Joseph Weizenbaum in.

1966. Who created Eliza the, therapy. Program, it's, a question and answer user. Conversational. User interface, and. Very. Difficult and problematic in ways that cause Joseph, Weizenbaum to, rethink his own total. Engagement with, computation, and artificial, intelligence, so. Everybody, knows that this is difficult to do and yet we continual, continually. Try to do it anyway. Nicholas. Negroponte II and Marvin. Minsky were friends, throughout. Their. Entire throughout. Nicholas's, entire professional, life until, Marvin's. Death a couple of years ago and here, you see a. Robotic. Arm stacking, blocks this is a lot, of AI problems, were, developed. In what we're called micro worlds right so you focus in on some small. Part. Of a project some small issue right, you're looking at edge finding, you're looking at computer, vision in fact you're still today, thinking, about computer vision in some very specific small ways I, know Justin was just taking a computer vision class at Stanford, and. You. Tried to throw away the rest of the noise and just focus in on that problem and that's. Appropriate for, some. Period of time but after awhile microworlds. Didn't. Actually scale, up and this gets to be a problem. But, as I mentioned the architecture. Machine group worked in micro worlds as well and ai. Ai. Interfaces. And this is a project that they did in. 1974. The software, show was a museum. Show at the Jewish Museum and. Information. Technology, it's new meaning for art and so here you see all of these cubes, isn't it funny like generator, has cubes and, urban 5 has cubes more cubes and, you've. Got this, set of mirrored blocks 400, married blocks of 5 foot by 8 foot pen and, a. Gerbil. Horn a, gerbil. Colony, living. There and it's gonna be difficult to read this but it says gerbils match wits with computer, built environment. And. What. Seek did was stack blocks and what, gerbils did was move them around and knock them over. This. Is the gatefold of the software. Catalog. The life in a computerized, environment and, there's, this great quote from Ted. Nelson who, says our bodies our hardware our behavior software, and it's. Paul pangoro, who made the contemporary, color, coin of mobiles and a, member, of the architecture, machine group who told me, it. Was a little bit too true because seek, tended to kill the gerbils in. Later. Days the architecture, machine group turned to command and control interfaces, that were explicitly. That. Had explicit, military connections. And applications. And this is because the only way passed. 1974. That, you could get work in artificial, intelligence funded. Was, to align it with command. And control questions. And direct, tactical. Military, applications. This is due to some, legislative. Things that, had happened in the 70s in the wake of the Vietnam War and so. This is the Aspen movie map it, is kind, of what it looks like Google Street View that you experience sitting over in a room in a chair you zoom down the streets of Aspen Colorado, in, an, Eames lounge chair, with. Joypads. And a. Video disc, produces. The images, for you they had a prototype video disc player in. The 1970s. And the way you get the images is everybody, goes to Aspen on vacation, and you load up the Jeep and that's, what you have right here, so. You zoom down the streets and there are other there, are other photos, that show, what's in front of the the. To smaller screens are usually, maps and are on touchscreens. I want to point out that part of the reason this project was funded was. The, it was in the wake of the. The. Jet. Rescue the the hostage, rescue in, Entebbe, Uganda, a. Number, of Israelis, had an Israeli, jet had been. Had. Been hijacked, and. The. Rescue. Of the hijacking, was, performed. By building a model desert, or a model, Airport in the Negev desert and prayer, death. Practicing. The rescue. Under, cover of darkness. In fact, this. I, think this is the subject now of a movie that's about to, come out or as coming out later this year and. So, the question here became what could be what, would happen if you were able to actually simulate and do, remote moving, and, move. Down these streets so, that would be the stipulation. For this project. In. Later years they were really explicit, about this and some of their the some, of the reports they wrote Minsky. And his colleague Richard Bolt wrote that we are proposing to develop human computer interfaces, on the one hand is sophisticated in, conception, as a cockpit and on, the other hand is operationally. Simple as a TV from either perspective the objective, is the same supreme.

Usability. Later. On they. Started delving into. Virtual. Reality in certain ways and questions. Of augmented, reality so. This, provocation. And mapping by herself is a. Postcard. Held up in front of the Queen's palace in Stockholm and imagining. What it would be like to have a screen that overlaid, information, upon the world on, the. Other side is, Westinghouse. Mapping, a Westinghouse, window, basically. A very, early. IPad. Like thing that. Was, still connected, to various, technologies, and you can kind of see in the background there's a Selectric, typewriter, and, a dictaphone. But. This was a mapping. Window and intended. To substitute, for maps in the battlefield, it had two dimensional, and two, and a half d. Capabilities. Depending on how you would flip it and this is apparently the first example, of a layered digital, map. I. Like. The fact that Nicholas, Negroponte has. Always understood what. He. Has been working on and the conundrums, and, contradictions. About it and in. His 1975. Book soft architecture, machines he said I strongly, believe that it is very important, to play with these ideas scientifically. And explore. Applications, of machine learning that tottered between being unimaginably. Oppressive and, unbelievably. Exciting. And. I. Find myself as as I look at these historical. Examples. And where we are today I find myself wondering if we need new cliches, because. When I do Google Image Search and. Pop in AI this is what I get. There's. Always this you know the wavy or. The. Behold, the hand and it's always that color blue it's like the the blue of the projector, when no one's in the room. Chaisson, lady I think, there no this is jQuery right yeah, yeah. Cyborg, lady. If. These are our views of AI, in the future I think we're in trouble but. I appreciate some, of the subversive, ways that we can look at this and some of the fun ways that we can as well. Is. Anyone here a fan of Jenelle Shane in AI weirdness, I. Nerded. Out so hard at i/o this year when I got to meet her she, she, does very funny things with AI that I'll talk about but I really appreciate that, she understands, what she's doing and she said you know it's life plays by the rule engine recognition, works well but, as soon as people are sheep do something unexpected the, algorithms, show their weaknesses.

So, Are these orange flowers in a field or sheep because she colored them orange it's not, sure anymore. Jenelle, has also done things like this, is 7000. Paint. Colors, fed, to a neural. Net and. It. Begins to, create, new colors all of which are terrible, and color. Names. Exactly. Yeah I've. Been reduced to tears multiple, times, Carrington. Burble, simp turd. Ly. Ranching. Blue thunder. She, fed the names to. The. Names of guinea pigs for the current Portland guinea pig rescue and came up with some names for guinea pigs. Here, we have hanger, Dan and, princess. Powell. And. My husband has said you know I think any kid could come up with princess Powell but it takes a neural net to come up with hanger dan, only. Meets machine learning can produce it and. And. Again when I think about these kinds of things I also find myself thinking about this, which is actually, it it apparently, was not Marcel, Duchamp who delivered, this to the. Art Show in 1917. This urinal, as his. Object but rather a woman whose. Name, I can't remember they've, just surfaced, that it wasn't him but this, was part of dadaism, where the, world was going to hell it. Was World War one and things were really, really dark and so, what does art mean when things are really dark means. That you've got to turn to absurdity, you've got to turn it upside down and, so art starts to look different, or. You look at someone like Eugene Ionesco who wrote the play rhinoceros. In. 1959. It's an absurdist, play and it's, where, a group of people in a French town bit by bit turned into rhinoceroses. And of course he's talking about totalitarianism. UNESCO. Wrote a play called the Bald Soprano because, he had started to try to learn English he's, a French, Romanian, and. In. Order to learn English, he would type phrase, after phrase from this kind of English book John and Sally go to the, store over, and over and he began to question the nature of these meanings it's, almost like he became his own algorithm. In order to learn this and he, produced, a very funny play called the Bald Soprano about. It or. You could even look at something like, boon, yell Louie baloon boon yell who did, Chien. Andalou in the, 1920s. 19 teens but. In 1970. 71 something, like that he did the discrete charm of the bourgeoisie which, is about a group of people trying to have dinner and they just keep getting thwarted, and it's, a content well somewhat contemporary, surrealist film but all of these things are ways to when, the world is. On its head, you upend it in art and I want to suggest that maybe we should be doing the same things more, frequently with our algorithms. Can. Anyone name what this is. Thank. You. Okay. We're saying that but yeah, yeah it's exactly what it is and, the.

Uncanny Valley is an idea from 1974. Masahiro, mori and what, I'm struck this, is the the idea that when robots are too similar to us they freak us out and he. Describes the eeriness, that that produces but, he, says something else in his piece in in, the article, he wrote in 1970. Which is that it's about how we understand, what makes us human, we. Should begin to build an accurate map of the uncanny valley so that through robotics research we can understand, what makes us human. Until. Recently Manuela Veloso was the head of the, machine, learning program, at Carnegie Mellon and I, was, at a conference that. She was speaking at and she was talking about machine, learning is successful, when it does what you expect it to do and for. Cedric price we, saw the, goodness, of when something, doesn't expect do, what you expected, to do but. There are some dark things to, the. Stories that came out earlier this spring about pollen tears predictive. Predictive. Policing in, North in New Orleans and. The. Fact that that, had, not been known not, even to, most. People in the city of northern New Orleans or city officials, I. Like. The work a lot of memmio knew aha some. Of you may know her work she does the project called uh missing. Datasets and you. Can't you can't parse what you don't collect and so, if you don't have the data you can't begin to parse it she, does this both as a tech, piece and a. Critical, piece and an art piece so here, you see a lot. Of folders. Full of missing, data sets and of course if you go look through them they're all empty. She. Keeps a github, repository, where. She, has. These missing, data sets and, this. One civilians, killed and encounters, with police or law enforcement agencies, is no, longer a missing data set it's been collected, and. So, she'd. Like to see more, of those and you know case in point sometimes there's. An activist, angle, to her work but, there's also. There's. Also work, that happens. As a result a group of Asian. Actors, on Broadway. Contacted. Her about the, roles that are available for, Asians. On Broadway, and that. Blue line is. The. Asians the, Asian roles in the play the King and I and. The. Other bright. Light blue that you see our Asian roles elsewhere, in the. 2014-2015. Season. Similarly. You can see the problems for black, actors or Hispanic, actors just, not a lot, available, at. That point but by collecting, this data it, became possible to change the equation and now, there are more Asian, actors on Broadway this. Is not a matter of an algorithm but it's just quite simply a matter of, the. Collection, of. Data and collecting. The data and naming it making. Changes. Oprah. But. I'm struck, by the things that machines don't do that we don't have good machines for picking raspberries and, we, kind of do for picking strawberries but, it's. Very very difficult to, get this right to manipulate, to do, this properly and I'm. Struck by the fact that if you I. Think. It, was in Sweden there was a mining, operation that. Had a truck that. Had some very nice. Machine. Learning algorithms. Operating, it more efficiently, and it was so efficient that it dug itself into a rut and couldn't get back out so. They had to introduce noise into the, system to allow for, the. Our. Love for the vehicle to not to expertly, dig itself into, the, ground. So. In closing I'll say that AI is new. But. It's also old and that. It is a matter of architecture. It's. A matter of infrastructure, and it's. A matter of how we design. This. Is something I see in my neighborhood, in Pittsburgh I go running by Carnegie robotics. Not. Totally sure how I feel about it. Yeah. I'm still not sure how I feel about it but, I think it's pretty compelling. But. This is the thing that you do come back to over and over again which. Is our man Cedric price who says technology is the answer but. What was the question. Thank. You. Hi. My. Name is Davide Malina I'm an analyst, here at Google and, I've. Been getting really interested in some. Of the things, happening on the edge of this space like generative design and some of this stuff that Autodesk, is working, on yeah and I'm. Curious. As to how. Where. You see the space evolving. Sort. Of like fastest, because you, know like 3d printing you, know there's usually industries, that pick it up first like where do you see like the first sort of mass-market case. Of architectural. You, know AI or intelligent it's really happening that's, a really good question and.

I'm Not sure if I if. I have a well-thought-out answer, but. I'd point out that the history of 3d printing is already, quite. Old and so what we see the uptake in now at. A, consumer, level or at a professional. Level. Well. And the other thing is I don't think that 3d printers, I think. They became, something that it, was said they were going to transform, everything in the world and so then everybody started getting maker, spaces and, I'm. Not sure that that has actually, printing. Objects has made the world changed but, the. Question of generative design, I think that architectural. II the primary, focus. Has been on form and representation. You know you you use, any. Number of authoring, programs and sign programs to produce. Produce. Work that you wouldn't be able to otherwise I, get. Very curious again about these old ideas that I still think are. Vital. About how they. Change, the. Design process and, the. Ingredients. That go in that make us think of something that isn't necessarily, just, ending. Up in form but ending. Up in how we construe. Design. But. That's yeah. Thanks. It's a good question now I'm gonna try and figure out the answer. So. Rohit. Says. Any thoughts. On bill Mitchell's roll, city. Of bits and or, urban, design on software, yeah, other, than Sim City Oh Bill. Mitchell. Bill. Mitchell was. At. Points, the Dean of the School of Architecture at, MIT here and the smart cities lab absolutely. Vital individual, he's actually a alum. Of my master's program and he died about five years ago and. I didn't write about him very much in my, book I was focusing on a slightly earlier, era. But. He's. He's, a totally, vital figure, and there's a woman named Ann Marie Brennan right now, she's. A graduate of my ph.d program in. Melbourne. Australia who's, been working on some, of his, work, and using his archive there so I think we're gonna see some stuff, forthcoming about, Bill Mitchell but. You also see his work in you. See his how he's inspired people including, like Anthony, Townsend, who wrote the book smart cities, and. Has, been thinking a lot for, the last 20 years on urban planning smart cities and development, I. Guess. From a smart city perspective, I keep going back to two. Big, concerns that smart, cities tend to be. And. Again I get to say this because I'm an academic I guess but there. It. Seems to be viewed as the locus, of really great big enterprise, deployments. Like if we used to get excited about enterprises. The scale of the building or, a campus. This. Is even. Yet still more tech. And a. Lot of it hasn't been very graciously. Designed I also think of that idea of microworlds, that the, question of scale is really vital but, when we say scale it means a lot of different things like is it 1 to a billion is it tiny.

To Great big. It has. Many more users as, it has many more stakeholders and, I think designers, and architects are really well positioned to talk about issues of scale. But. It's it's, the hardest thing I think for us to figure out what we mean when we mean deploying. On that kind of scale Thank. You Rohit. Hi. My name is Christine, I. Just. Wanted to ask like what. Are you most excited about. That's. Getting like research now or, an. Idea that's, you. Might have just heard about like what in this field are, you most excited about, thanks. For that question that's a really nice thing to be asked. Well. I recently received, this named professorship. And it's. It's. The, Kane L Gates professors, associate. Professorship and ethics and computational, technologies, I'm, not an ethicist I'm, a designer and I'm a historian and, I'm. Really curious about what, we, talk about when we talk about ethics, because I don't think, that it means. Ethics. Per se, I. Like. You can't just sprinkle some ethics on top and make the food salty, and and, that, that doesn't work and it makes me remember that time in like the 90s and 2000's, when people discovered, usability, and said, we need some usability on this website and and. Like no you don't want usability. You want it to work you want, it to be good and I'm. Gonna leave here and go to Stanford, to meet up with Fred Turner. Who. If you haven't read his. Recent. Pieces in logic, and in, 0:32. C you. Should the 0:32, C piece came out this week. The. Piece and logic has a few months old so he's, the head of the the. Communications. Department he's, a historian, of, among. Other things Silicon, Valley, and. He. Were, grappling with a question, about education and. He's. He's speaking about engineering. Writing about engineering education but, I mean I think anyone, who's gonna work with stuff how how. Do we educate. People engineers. Are increasingly, Fred, will argues. Oriented, and then anything else is relegated to. Other departments, but. How, do you make an ethical, approach not be a blow-off class, how. Do you make. It be a part of the data sets that your, the problem sets you're working on. Part. Of my, interest in casting, the role of it of design. To. Work in different kind of areas is. Because, of that I'd like to, you. Can you can work with designers on framing a problem. Determining, what data should be collected how, it should be used how it should be visualized, or explained. So. I have questions about that I have questions about the. Possibilities. And impossibilities. Of, explainable, AI and, honestly. If you just sit me down with like a new history. Or a bunch of documents, about the history of AI you won't see me for a while because I'll be very happy if I can go into. Into. That stuff but I think we have some very human questions, that, also meet the, jobs of the people in this room and the people on my campus, and probably the places that we all studied. That. Need broader approaches, than just the, most efficacious and, effective, and efficient, approach. Thanks. So. One more question from the dory also. From rohit. Why. Aren't there more real. Architects. Celebrated. In the field of software architecture other. Than Christopher Alexander, bless you for asking this is so great, thank you um yeah. I asked, Alan Cooper a, couple, of years ago can you name another architect. And, I. Did this to Kent back to and, silence. Fir and I. Think that Kent, Beck said lacor BCA and I think Alan. Cooper said I. Was, at John Portman the guy who did the Embarcadero. Center and, you, know the. The Bonaventure, in in LA. It. Would be better if we had other architects. But I think the other thing is that Christopher. Alexander. As a, mathematician. And. As something, I mean it's something that everyone in the room deals with you work on some very tiny part of an absolutely, huge problem, how, do you go back and forth how, do you keep that in your head and Christopher. Alexander, can explain, that to people in. A way that is really, really that really, resonates, with them the the working, on these two scales and then, explaining. Why it. Actually matters on like a cosmological, sense, as light stuff is is very woowoo, and, and. That that really resonates, and it really, doesn't resonate with, most, architects, to build buildings, so. I. I. Would like to see Chris or I'd like to see Cedric price get taken up by everyone else because he's fun. Alexander's. Not much fun, but. Cedric's have a blast, and.

To, Get, to know other architects. Out there, I think that the classes, of people over, about the last four or five years to graduate from architecture, school, or. Design school or those of you are in the spaces that crossover really, doing interesting work and interesting, ways of thinking about it if. You want to look at something fun look at my friend Fred Sharman's, work he's an architect he's a professor at Morgan State and he. Has a book coming out on. Space. Architecture like, space. Exploration architectural. History's next year so fantastic. Stuff. Throw. Your talk I think. Sometimes. When you mention AI like. I would, consider, it as like, more like human-computer. Interaction. Mm-hmm, the reason I'm saying that is because, like. I'm, an engineer so when I get my PhD like, years. Ago so, like. If you have a PhD in AI is, not, as easy, to find a job as now yep, right, so in. My understanding like like we have a lot of breakthrough, in AI like core, technology I'm talking about like for example of a gold yeah or like image. Recognition totally. IBM. Watson, right so when they won't, the chapatti show, at that time yep, so we're talking about like they're a different type. Of AI, core, technology, advancement. And, that's recently bring back the. Tran of AI, but. This. Is quite different from. What you're talking about like the human computer in action so and I think IBM call, it like cognitive computing, so. Would. You would. You agree with me if like I try, to understand, when, you say AI is, more like human, interaction. Which is not Cori I ever from that we're talking about so. Yes. And no. Thank. You for that that's great and to. Be to, be clear the focus, of my research in. This book is, 1960. I mean. To some extent 1948. Because, I look at cybernetics, a lot but. To. Really. About 1977. 1980. The. Parts, of the book that are focused, on on later, material, are not. Focused. On questions of AI and, I. Agree. With you that they are questions, of human-computer, interaction. But. I, would. Say that. When. Negroponte II was writing this. Book he was writing. The book in 1970. And doing the, work in 1970, he. Was developing. Interfaces. With. The AI lab, using. Their, technologies, and their environments, and basically. Tinkering, and working in in, that capacity, so, yes they are its input and output devices and. And. All, kinds of ways that the lab over, that that period of time played with that it. Becomes, human-computer, interaction. But I don't think HCI, is the field really kind, of comes together, in.

That, Term. Until. The 1980s. I, I'd. Need to look at my notes to remember, exactly when, I will. Say that I'm very curious about what happens in the, HCI community, around, 1990. Because. There's some interesting crossovers. That I think have. Have. Things to do with. HCI, and, interaction, design in, ways that are in. Things that are really vital. And. Yeah. I would also agree with you that things. Are way, more complex, today and I. Have skated, by all, of it. So. My. Understanding, of it is more nuanced, but again I'm also a. Historian. Of technology, and architecture and, I'm. Not someone. Who got my PhD in, AI. But. Yet when you start looking at it today and the. Implications. And. The. Different kind of terms. That people are claiming cognitive. Computing deep. Learning. There's. A whole, different, set of things to start looking at and studying their, their histories, their going back to the 1980s and 1970s, that. I haven't yet gotten into, but. It's you're you're right and and, yes. But no but yes. Thank. You for that. Thank. You so much everyone, for coming out today and thank you Molly for your presence here thank you. You.

2018-09-19 17:58

Show Video


Christopher Alexander is not enthusiastically embraced by architects and planners is probably due to his overselling of the pattern language. His aims were much larger than combinatorial flexibility and modularity. He promised a new urban theory. His 253 patterns were inadequate in the face of modern urbanization with mixed uses and complex built forms. In practice, the pattern language fell short. In theory, the pattern language was inadequate. Hence the disappointment. In fact this disappointment became apparent and Alexander himself noted in the very first project he undertook for one of the neighborhood plans in San Francisco. A similar disappointment came about in computer languages that mimicked patterns in object oriented languages (I'm mainly thinking about gang of four). Patterns fell woefully short when expanding methods and classes. Large-scale software projects became unmanageable and a crisis ensued. Functional languages, such as Haskell, solved some of these problems in earnest for the first time, and a major crisis averted. Now older imperative languages such as C and Java are retrofitting functional programming concepts such as lambda calculus, lazy evaluation, monads, and so on. Patterns in software, like in the brick and mortar world, are no longer seen as the panacea for modularity. After decades of false starts, we have realized that modularity is not about finding connections in the superficial methods or properties or features. It goes much deeper to the very the functional aspects. In software, it is found in deeply embedded mathematical operations. In architecture, it is found in deeply embedded aesthetics.

This was a nice tour through the systems and ideas people developed in attempts to create useful automatic machines. And it's somewhat interesting to see people having these ideas before there was any technological foundation to support it. But there's no depth in this talk, it's just a organized collection of fun facts.

brrrr... no children, no cats, no proper tasks - just bs as result and without respect and epathy to others. but a lot of blablabla


funny challenging. You.

Other news