13:27:09 Well, all right. so let me start recording I have a couple of short announcements, and i'll introduce you. 13:27:26 Okay, I'm delighted today that We have bill Buckstone as our guest lecturer. 13:27:35 A couple of quick announcements we're hoping to have homework once in this grading today on our 2 starts Thursday. 13:27:41 I mean of our research on what? Yeah, Homework? 3 starts on Thursday. if people are interested in sending me some menu examples, you can do that today or tomorrow that a few in already some pretty cool ones? 13:27:53 Everybody's doing Super well on homework one but we'll get to those as soon as we can, and and i'm really delighted that Bill Boxen is able to we are first the guest lecturer a 13:28:08 few rules according to the university. it's okay if you ask questions, and your voice is captured as long as you agreed to app that it will be captured. 13:28:21 And but publicly on the web with the people who are on zoom, however, cannot ask questions or verbally, because then your video will show up, and that's not allowed. 13:28:30 So please go ahead and ask your questions. If you chat them torn, or I will ask them. 13:28:39 And Bill can kind of see us in a little tiny window from the video like normal. 13:28:44 But i'll try and keep look out for people to raise your hands. 13:28:49 We definitely would like some questions as we go along. we have, you know, the usual time, or will be 2 40. 13:29:01 I like normal, and Bill is in Tranco, which is on Eastern time. 13:29:08 So with that well and i've particularly delighted to have bill speaking with us, he in this early in his career, invented a bunch of interested in gestural interaction techniques. 13:29:25 He created some of the i'm not saying that trans Oh, not on Yeah, no, it's fun. 13:29:44 He. He also did a bunch of work on new interaction techniques when he was working for alias, since it was autocad 13:30:10 He also did a bunch of really interesting work on tools to help people build user interfaces. and, he's been a consultant to many companies. 13:30:22 We currently for Microsoft and he also was my own pizza. 13:30:32 I've known bill for a really long time and I delighted the teeth willing to give us overview of part of his specialty, which is 2 hands working with 2, All right, ready to go okay and just be clear if it's for most of you that are in the room and on camera just raise your hands if you do have a question 13:31:12 or send a chat to brad and he'll put up his hand, and i'll take them I don't mind being interrupted. 13:31:26 I'm happy to have a conversation and we've got lots of time. so 13:31:38 But we got 2 hands and it's interesting because they're both pretty useful, but sometimes they write together the new plan. 13:31:57 The question. That sort of undermines a lot of the thinking. 13:32:04 Here is is Is that one task? or 2 tasks because i'm moving 2 halves, moving together all the third task may be thinking about coordinating them, But you think about them it's just i'm clapping my 13:32:19 hands. And now you're thinking, you just came to this pretty good university. You're sitting around in a class, and somebody's just telling you. 13:32:34 Oh, you can clap your hands. But that's a really interesting thing. 13:32:43 What You're doing there and I want to talk a bit about how interesting that is. 13:32:49 So now it's going to get stranded because i'm going to mention the polling board part, and I use him for a metaphor, because it's a pretty famous pose that he uses when he's there around the time when brad was a grad student with me and so on and so forth my kids were little they're all going up and 13:33:09 married that time. They say hit that. Why are you going all over the world and giving talks? 13:33:22 My answer would be because i'm a genius and They said, But, Dad, you're not a genius. you're an idiot. 13:33:34 I mean that would if they were being polite mit and they said. but dad, everybody knows we have 2 hands and they know how to use them. 13:34:00 Then I would say, Well, if you look at the working with computers, you'll never know it except for one thing when you're typing, which is something which you apparently can do with one hand, So when you work with the computer you use one hand, to do 13:34:12 you. you use 2 hands to do what you can do in the real world with one hand with a pencil. 13:34:20 The rest of the time mit ctl and you're at that. So what you really are is you're like Napoleon Bonaparte for, and you've got these fancy tools and I just sort of say what the heck is that all about that makes no 13:34:42 sense. Now the reason I know I have 2 hands is because I spent 20 years in my life as a professional musician, so I played a number of instruments, but saxophone was my main. it's been but I know I can work 2 hands. 13:34:58 Together, and and so I brought that genius such as it was and of course i'm being festivious to when I started to work with computing, and the only reason I got into computers. 13:35:20 So I want to start with things that are a little bit further away to just get this concept going. 13:35:28 And what were we heading? This isn't my car it wasn't my car? I had a blue one of these except mine wasn't shiny when there's an undergraduate built one good one out of them sold the rest for parts kept a spare engine and so I had for $120 my whole 13:35:54 time in the sixtys that was pretty cool now the thing about the sports card is, it had a stick shift, and I was an undergraduate, and so I'm i'm going off and imagine i'm driving down the 13:36:14 road, and my girlfriend is sitting beside me and she says, Hey, Bill, this is really cool. 13:36:22 There's a stick shift? can I change the gears and how can I refuse? 13:36:30 I mean really, how could I so i'll have to change your gears? 13:36:36 So that's all It is just change the gears and Furthermore, 13:36:52 I also know that there's actually 4 things you have to operate to change the gears on a standard transmission. 13:37:04 That's So you have to steer which involves one device which is the steering wheel you have to work the throttle which you and of course you have to change the gears. 13:37:14 So that's actually a lot of stuff that you have to do. and if ever you, any of you who drive a standard, you'll know that. 13:37:15 So. because i'm. a genius I knew if I could remove one of those tasks. my task would be easier and make it do that one simple task. Don't do that you will do that. once and you will never do it again. 13:37:42 Because it's my car I had to fix it and you kind of leave a Trans. 13:37:45 It didn't work, and and so you could think about why? 13:37:49 Because if we do a task analysis, we'll see that steering Okay, I got to kick the car moving when I'm changing gears. 13:37:58 Otherwise you Can't use it. I have to pop the throttle to get to get things rubbed up. 13:38:06 I have to depress the clutch and later. on Release it mit ctl, and and you have to move this stick which you have to grab It's a you know, device acquisition task. 13:38:19 Then, so I use 4 sub tasks, each of which he constructs, and even more complicated is All of these things have to be done in parallel. This stuff is done sequentially. 13:38:34 The So here's where the I was going to say the transmission falls out of the car. kind of thing. 13:38:50 If you don't get that there's just going to be the most horrible grinding noise, because all of those things have to be synchronized. 13:38:54 And so the one or 2 things here is changing gears, something that we conceive of like clapping hands as a single task actually is a compound task, consisting of a lot of secondary tasks. 13:39:11 When you are skilled they completely are aggregated together, and aggregation is one of the teams used in psychology to describe skill acquisition. 13:39:25 You get a number of small skills and and so gear changing it's just a single thing. you do it in one gesture. 13:39:34 When you are skilled it's automatic you've tried to ski. 13:39:43 It's the same thing. How do you coordinate bending your knees waiting the floor Ski So was an interesting lesson now. 13:39:52 All of that led to something that happened when a woman named Maureen Stone, who was a well known in our field, who was worked with me at Xerox Park. 13:40:02 Later on, when I was there, came and told me bill you do all these neat things sort of interesting, but they don't make any sense. 13:40:13 There's no coordination. so you've just you don't have any program, and I knew she was wrong. 13:40:21 But I didn't have an answer it's really important mit ctl, and that when people are candid with you, and you have to be the same way when you're talking brad if he doesn i'm gonna say thank you tell me what because you have to be 13:40:35 willing to rather learn something. Be wrong so iran correct me rather than walk away and feel youth lost face, because you called me out in public, or you. 13:40:47 You're embarrassed to tell me i'm wrong, and therefore i' walk around diluted or I did that. 13:40:51 And so I wrote a paper called Chocolate or Phrasing, and And and this is critical aggregation of sub skills is the core of skill, acquisition, and second, of all the way you get accelerate cementing these sub skills 13:41:00 together is using in my experience something I call kinesthetic continuity. The catalyst to glue the sub skills together is condu. 13:41:06 Obviously, it happens through the power law of practice. But the way you accelerate, minimize the practice because the power laws are expensive is how do you phrase things together like when you're playing a note? 13:41:11 You know, from breathing, and you know from a hand gesture when you start and finish. 13:41:13 So you know the beginning end, and these things just the breath or those things tie things together, and it's usually due to motion or tension, and those are your frames. 13:41:20 When you're trying to, from the kinesthetic perspective phrase things together, so that things are seen as a single gestalt rather than a bunch of planning of subtasks where you have to plan and you have syntactic errors and all 13:41:27 sorts, and you make mistakes. They can be parallel or sequential, or some combination mit ctl. And these are things we know inherently music. 13:41:34 We do this all the time, but we don't get taught this generally, and then the Slide deck, which i'll share with you the paper that was my mentor, and this is a woman named Phyllis Reasoner she published a paper in 19 13:41:38 81. All the papers and cognitive modeling If you've taken a red any cures of pulse and pain in green. the comrades command language grammar for me. 13:41:45 This is the best paper, because she doesn't stop at the device level. 13:41:48 Everybody else says there's all this cognitive stuff and Then There's the device level that's something else. we're not gonna deal with that. 13:42:15 No, no! she talks about the Kenneth the whole thing at the surface, so that you can actually predict errors, learning time and speed of execution, and she does it through very simple 3 heuristics about of a bnf 13:42:24 grammar. You can read the paper, but the basic thing is, the more terminals, the longer the productions and the more productions, the more longer can take. 13:42:33 And and and this longer will take to do okay so and then there's a chunk kind of phrasing paper. 13:42:40 So question, Mark for those of you. I can see how many write with your right hand. 13:42:47 How many right with your left hand? no left-handed people in there? 13:42:52 Wow! doesn't seem like it okay? who did not raise their hand Okay, it's too bad because a person who didn't raise their hand is the one who's correct you're all wrong you think You're right? with your right? 13:43:14 Hand. But you don't your right hands involved but you don't, which is why I set you up earlier. 13:43:23 When I said, The one thing you do with the computer with a pen is is, is right, or without a with a pen, you can write with one hand. 13:43:33 Actually you can't But I was trying to mess with you so what is that? 13:43:39 I'm gonna play a video this is just me reproducing, hey? 13:43:47 An experiment that a friend of mine, and if Gyard did and he was introduced. 13:43:57 His work was introduced to me by one of my graduate students Cockabash, and when I read the paper that Paul gave me, I literally got on online, found out where he was giving a talk or where he was who he 13:44:10 was, I got on a plane and flew within a couple days to the Southern France, where he is giving a presentation, walked up to him and said, You have no idea who I am. 13:44:27 You have no idea what profession I work in but i'm gonna make you famous because you're gonna be famous in my community the minute they know about you. 13:44:32 And it happened I was right, but it wasn't because of me I just made the introduction. 13:44:39 I just brought them together, and this is like this because this is how elegant this man's work was so we tend i'm writing I'm just but i'm not pretend i've just finished writing my notes for a thing 13:44:51 for, and i'm finished so here the arts on this 8 and a half by 11 sheet, all nicely lined out terrible handwriting. 13:44:59 But we think we understand what i wrote But this isn't what I wrote. So have a friend, Eve Kyard, and he did this study, and this is a replication of what he did carbon paper on the bottom of the writing 13:45:09 surface. so the sheet of paper hair. So that is actually what the carbon paper said. 13:45:14 I wrote, and what you'll notice when you compare the 2 this is way smaller. 13:45:20 It's narrower and it's rotated and What this clearly shows is that while he was writing I was nicoting a paper rotating it, and signed it left and right as well as up and down so that the zone 13:45:42 of this is works all the time. What this says is the gods in the detail. 13:45:44 Other. The writing is a bimanual action that requires this coordinated skill of the 2 hands using different functions. 13:45:58 Touch and stylus, if you will, gives us a hold of an insight on how we write with a pen it on any endeavor that we set out from writing or things much more complicated where we want to support batteries interface, we have to understand the activity the intent and the human capacity at this level whether a 13:46:19 capac. right? So this is you can sort of say well okay That's writing paper. 13:46:29 There's there's no computers there but I would like you one of the reasons that pen computing is at such a pathetically poor state. and it's one of the examples where how, the power of the potential of the hardware is so far ahead of the 13:46:49 software. , is because the software and the hardware don't understand people that if you're trying to draw or write or keep treat your pen based computer as a piece of paper, it Better have those properties So you can move it. around because you you do not write from the top of the screen coming 13:47:18 down and over the pessel. you really have a very small zone of comfort, and you move the thing around, and you have to find a way to do that. It's by manual. and there's virtually no system. 13:47:38 That does that unless it you as you x people you want something to invent and work and on, you can fix that in software. if you build interactions and you use b manual techniques. so on. 13:47:49 And so forth. I'll leave that to you but what I could say is when I read that paper in Southern example, and the theory is that we're behind it. 13:47:56 Man I knew I had to meet this guy and he had no idea it had to do papers. 13:48:02 So let me give you an example just to point out because i'm going to tell you a couple rules here that he came out. 13:48:13 But there's there's some women in the room and and by the way, i'm not being sexist i'm just telling you something that's hey? 13:48:29 Women. Do you know why the buttons on your blouses are backwards? 13:48:33 They are wrong, especially for you and men. Do you understand? why yours, are right, and why the women are wrong for them. Because i'll tell you why one of the consequences of Gillard state is but give it the rules and it's and this all comes out of something called the kinematic Chain model by manual 13:48:56 asymmetric packs in a motor action. 13:49:02 Okay, Okay, let's get through 2 hands by manual kinematic chain means things are linked together. 13:49:08 The movements are linked together. but asymmetric. this is symmetric by manual action, or so is this 13:49:20 Either it's that way, or that way but anything else where you move in different directions. 13:49:25 That's asymmetric. Okay, and motor action is just because i'm moving it's it's the mechanical part of my skeleton all right? 13:49:35 So the assets sort of theory one the non-dominant hand, not the left non-dominant hand moves first. 13:49:47 The non-dominant hand sets the location of the action. 13:49:52 The dominant hand moves second the dominant hand desertified motor action here's what this has to do with blacks. 13:50:05 And why they're wrong for women here I am i'm wearing a men's sweater. 13:50:08 A man here's my hand first step nod dominant i'm, right-handed, left-hand Bruce in left hand positions. 13:50:15 The buttonhole holds it in position, the dominant hand comes, it gets the button, and then it does. 13:50:26 The fine motor action at the location defined by the not dominant hand, and puts the button through. 13:50:34 And since almost every action is a a by manual action. is a is is an ace follows the kinematic chain model. If you start looking at what you're doing, threading a needle follows exactly the same thing there's lots There are some 13:50:47 except the bat if you're designing anything for a computer. 13:50:49 It's by manual once you understand these rules you know you don't have to be able to do everything with both hands and the same assignments. 13:51:01 Other you can you don't you don't have to be wishy, wash. you say all that a 1,000 flowers blue. 13:51:05 No you can be in a trial safe if i'm dealing with a right-handed person. 13:51:16 You're gonna do this with your right hand that so only have to provide those functionalities. and I can tailor the individual devices separately for the properties of those 2 hands. and the functions that they're good for and Furthermore, we live in a right not everybody's right-handed So for accessibility you should make sure that you can 13:51:32 it' which, by the way, in arcade video games they used to have them. 13:51:36 So the track ball was in the middle. I need a redundant firing button on another side, and I could tell you, depending on the game where the most acuity was which control they used with what had and so this is all really interesting because it gives you guidance in terms of 13:51:55 how How? you don't have to do studies you can Trust the science and and notice exceptions, and then try and tease that out. 13:52:02 Non-dabin. hence first dominant, hence sets location down in hand. So $9 sets a location. Donald moves second, and a dominant hand visit by motor action. 13:52:18 And so this is the this is some foundations. these things recognized when he articulated them. and that's why keeping up in the literature this and I realized, help me see It's the same thing with bills reasons where it could help me. understand the things I was trying to think about and they put words and a model to it that. 13:52:39 I could build on and and it's this whole interaction of what's going on in the community, and and how the research and how ideas and innovation occurs. 13:52:47 So I want to step back because you could sort of say, Hey, Bill? 13:52:49 Not everybody. There had to be 2 handed stuff before. 13:52:51 It must say yes, absolutely I used to have hair that's me I was a musician that's 1971, and it's the first computer I ever saw, and I drove up, and the only reason I went there. 13:53:09 Is my step. Brother loaned me his Bmw. 600 Cc. 13:53:15 Motorcycle which I loved, and I could drive up from cakes in Ontario up to Ottawa to Google work there, and I worked from 6 in the evening till 6 in the morning. 13:53:19 Then drove back down to Toronto to do the film sound, to do the up, to compose the music for a film soundtrack, and all the music was composed, not and performed on that computer 100 and This is 1,970 13:53:32 one remember Macintosh came out in 8471. 13:53:35 The mouse was first shown at the mother of all Demos, generally shown, is 68. 13:53:42 This is only 3 years later, and this system had a mouse. 13:53:44 But also, if you look here, by the way, it even had a piano keyboard record play music into it. 13:53:52 It had a color graphics display there. the graphics display here, a according keyboard on the left, and a mouse and or a scroll wheels that I could work there audio tape recorder that I could record the music on right off the 13:54:08 bat, and you could synthesize things in real time. 13:54:11 1971 There's 3 places that we're doing amazing stuff. In those years Lincoln labs that part of Mit Angle Bart's work at Sri and this center there. 13:54:24 Nobody's heard of this place. but on computer animation that you see a pixar and stuff huge amount of that came from there has to be aci. That's why there's a lot of chaos and the skill so 13:54:45 here's bird song so clearly I was just that had two-handed input for someone in 1,969. 13:54:54 Oh, but I know his West Clark built this computer It's a first transistorized computer. it's called the Tx. 13:55:06 2 birds working on a bit later notice he's got a life pen in his non dot in hand. 13:55:15 Hey? it's gonna be left-handed and He's got the keyboard in his right hand. 13:55:32 I'm not sure. Why, this is scaling Okay, so Okay, but that's the stepping backmart the precedence to my work on by menu, but computers did have people working with 2 heads. So let's look forward. 13:55:59 Instead of backwards. what's the difference well first they were using compound tasks that's cool and pretty subtle. 13:56:12 But in all of those cases the non-dominant hand was pushing buttons and doing discrete actions. 13:56:20 They were pretty much in one place, and the dominant hand was doing all the movement come back to clapping. 13:56:28 Both hands are moving. Fact of having people using both heads each one working together, but with dynamically for continuous tasks. 13:56:46 I haven't been it had been done but it wasn't part of the it wasn't being done in any structured way. 13:56:54 And so here's a provocation this is What led up to the first paper that Brad and I wrote together. 13:57:12 Actually is it the only paper we actually co-authored the box of Myers paper? 13:57:14 I turn. It may be that's pretty scary hey? 13:57:16 We should work it again together. so it's this I do this like a kindergarten, because it represents kindergarten thinking back around 1,984 I've done a bunch of stuff and i'll show you examples of it where I'd already been doing this stuff and playing it 13:57:30 with, and I was speaking at Sitcoms, which is the big conference where everybody wants to get to. 13:57:35 If you want to get a good Kennedy or get a job and something that in a university you want to publish there. 13:57:39 So I was giving a talk, and I had the audacity to say, Hey, you know, like we should be able to. 13:57:52 We can do because we can do things with both hands, working simultaneously. 13:57:56 We do multiple Tasks. and all of the geniuses all of whom had Phds of verse brilliant, and they were, Brian said, No, no, no, you're wrong. 13:58:07 You don't understand psychology. you cannot do 2 things at once, because they'll be task interface because you can't tap your hands and rub your tummy. 13:58:13 Well, first of all, I right there in the State said, Well, I can do that. 13:58:15 I could walk in 2 gum as well. So I came back and said, How is it possible the really smart people mean? 13:58:27 Really people are respect don't understand and don't believe this is possible because they experience it every day, but because they experience every day they don't notice it. 13:58:39 The truth is hidden in plain view, but you're so used to it. 13:58:42 So wanted to do a study and this led to this paper, and that's Brad's head that's a Perk that he actually worked on before he came built at Cmu or for 3 rivers which is the outgrowth of 13:58:57 cmu wacom did chine the tablet little slider box that I, Fred guy for Dorco and I built from the papers here, and i'm gonna talk about that and that's sort of the key this 13:59:14 is the paper that i'm hanging around percy because this is fun, because Brad's to your profit, and but it's a really. 13:59:22 It was a really important paper for for me so here's what it was 2 studies. 13:59:28 And by the way, those of you have shall we say more sophistication experimental design? 13:59:33 This experiment could never get published today, because of our self. we say our dubious use of statistics. I think we could have got it if we said it's a single experiment with a and this that's put it this way it doesn't but but it was it. 13:59:57 Was neat it just how we described it. wasn so compound task, position, scaling, And so the exercise was. 14:00:01 You have a screen with this great big pattern on the front a little. 14:00:07 We think that's the same shape. on the bottom with the cross and both of them have a cross stairs. They both describe a rectangle, and you have 2 devices. 14:00:15 A slider box that has these things You can slide up and down there like a treadmill's and a tablet. And so the first thing happens. 14:00:22 Here is you. you you really try to do is get that both to overlap with that target. 14:00:29 This is the tracking symbol that follows the movement of the Puck on the tablet, and you and you want to get the cross here of this on top of the crosshair of the target. 14:00:38 So the target appears stationary. You must move out so what happens here? 14:00:41 Is there it goes That's what you do with the copies If you did, you see that? 14:00:45 Go so, and the sliders have to make because i've done. 14:00:52 Now is positioned the the tracking symbol over top of the target. 14:00:59 But the secondary task is to make sure use the slider to make it to scale the size, and it's in 2 dimensions with the origin of the center. 14:01:07 And so if you watch over here now, this is what the Slider does. 14:01:09 It. Just lets you grow the size into orthogonally from the position. 14:01:16 So it's a compound position scaling task and i'll just play a video. 14:01:25 And because the video those days was terrible that how and and it's gone through generations that because disk space men digitizing. 14:01:36 So for all kinds of reasons. it's good you saw this before you watch the video. 14:01:38 So there's The target there's the box and The first experiment involved placing one square the tracker on top of another square target. 14:02:08 The position of the trial was controlled by moving a Puck on a graphics tablet, using the right hand in order to complete the task. The size of the tracker must also be scaled to match that of a target this is managed by manipulating a continuous slider using the left 14:02:19 hand. when both the size of the position match the system signals completion it. 14:02:26 An acoustic beam. A second thief signals the start of the next trial. 14:02:35 Several trials formed by the subjects performance data is taken for a training for experiment. 14:03:03 One took place in 2 phases phase one we trained for the positioning task, but by doing a simple target acquisition caps. 14:03:13 This duplicated the actions in the final experiment, the only difference being there was no scaling involved. 14:03:19 The second phase of the training involved practice on the scaling tasks. 14:03:26 Here the position of the target stayed stationary, but its size grew and strong. 14:03:36 The subject then matched the growth, using the controller who would be used eventually in the experiment. 14:03:42 At no time we're both devices train for some cases nor the subjects ever informed that both could be used at the same time. it's entirely possible to perform the experimental task or completely sequentially that is position by aligning the crosshairs of the 2 boxes and then 14:03:57 scale in our training. We attempted to bias users to adopt this strategy. 14:04:06 Another strategy is to scale the tracker while this being dragged and target. 14:04:15 While we don't have to proper this particular. strategy we believe that users would nevertheless spontaneously use it. Our hypothesis was proved to be correct. 14:04:42 So the thing that's interesting about this and Why, this is interesting is, And this is an important lesson that transcends this example in all of our work. we actually every research in hey you biased your experiment in the design 2 nudge 14:05:07 people towards satisfied in your hypothesis. 14:05:15 Our strategy is always to nudge people away from our hypothesis, so that we can never be accused of the former, so that if we can put in clear biases towards in a training towards what we don't expect to 14:05:41 happen. What happens then is, if you do, if your hypothesis is in fact met. You have both. 14:05:54 It's all the more compelling and and it's and that's one of the reasons why don't do your study. 14:06:11 See what happened, and then formulate the hypothesis Okay, that's that's there's a the better. 14:06:20 You have a stronger hypothesis and with a good rationale As to why you'll get it the more you can bias against it, so that you can get these really strong results that are compelling intuitively as well as statistically and in this case it's good the intuitions were good because the statistics 14:06:36 weren we're we're actually interesting and the way. is just sort of giving us the things of the subjects, all of them, except for one, did things in whole or part in parallel. Remember, they never practiced it. 14:06:58 They were never told about it, and to make how clear that was because you have a script. 14:07:06 So they all get the same instructions. the one participant participant who didn't do it did it because mit ctl, and they intuited that we told them that we didn't want them to do it that way they intentionally 14:07:22 av So the it's really really interesting but the other. 14:07:38 There's 2 other parts here that business is that the performance. 14:07:42 The quality of the performance in terms of how long it took to do the past correlated up positively with the percentage of parallel activity. So this is the thing all we did in the papers. We just showed a scatter plot of the data. 14:08:12 We didn't take any measure despite this her thoughts of this experiment. 14:08:19 She actually did a Postdoc, with me she just laughed at this. and I didn't understand enough to realize why what was wrong? That is obvious. 14:08:33 Well, no, it's. so the point is that the one thing What was interesting here was just, but it proved all this study did is prove. 14:08:56 Yes, you can tap and rub your your head and rub your Tommy at the same time, and you can chew and come. 14:09:13 And now we're gonna do with what we convinced ourselves with that demonstration. 14:09:25 So we went to another study, which has far more to do with the real world functions that had a problem that Actually, you have even today 100, and and that's why I mit Ctl. 14:09:40 And I was thinking about what I want of all the things i've done what I wanted to talk about this is because this leave a lot to be desired relative to where they could and should be right. 14:09:50 Now, by the way, if you want to be good in this field don't expect to be happy i'm chief scientist of a company called Silicon Graphics, and the CEO sent to me once when I was whining he said they 14:10:07 all. he came from Texas bail Your job is to be frustrated. You are frustrated. 14:10:16 You aren't doing your job bill if you can't handle it. 14:10:21 Get another job. and he was right but it's not enough to be a wider. 14:10:36 I'm good at, what I do to the extent i'm good at what I do it's because I don't know how, and it's, and I refuse to take responsibility and i'm really good at laying the blame at the person who designed 14:10:46 it. but that makes me a wider the next step is and then able to say, And, by the way, here's how to fix it, and then volunteering to help do so. 14:11:00 If you can do that without pissing off the people who you're trying to help. 14:11:02 So you alienate them. Then you can have impact but it's a really difficult thing to do, but every time I'm frustrated. 14:11:04 So scrolling mit ctl and so something everybody I could this image could have come from windows or at Mac, or any word processor today. 14:11:09 There's panel wire here, there's text for some data Here you scroll up and down and you want to select something that's in you have to first navigate to it and then select it so this is a compound. 14:11:19 Navigation selection test, and to realize that these 2 things happen together all the time. 14:11:28 And if you have to move from Widget to widget with your mouse, how to state it. 14:11:31 You know this language from typing. You want each hand to be in home position for the task that is appropriate to be signed to it Mit do the scrolling navigation with the that hand and even home position of the mouse to do the 14:11:44 selection, because I got to point me here in the navigation here, because those that G. 14:11:50 Yard tells us that's the right thing to do in general So here's what we did little scroll pat here. I used to touch pad this time. 14:11:57 I'll tell you why and the the same pad in the same computer. So let's let's take a look at what goes on here. 14:12:05 There you get The document is a stylized document. Every thing On the left is the word left to the left column on the right column is the word right, and the middle column is the all it rows are numbered and there's sort of These diagonals to help you know that things are going up and down when you 14:12:28 scroll and all it is is the instructions are 36, right, 5 middle, and and you get that there's a queue. and then you have to get there and tap it. 14:12:41 And then it gives you the next. So here's the video just to put some meat on that particular phone. And again, this is Brad, you see. 14:12:43 Hope, having determined that users could handle 2 devices at once, were then entrusted in examining the how one-handed versus two-handed usage compared on a common type of transaction this was the basis of experiment 14:12:57 2 here. the user will select a word from within a document. 14:13:03 However, the target was never on screen. Therefore they first had to navigate to the appropriate location of documents, and then make selection in the one-handed technique which is being demonstrated Now navigation is done using scroll bars small 14:13:24 arrows in a manner model of bomb, the apple Macintosh computer. 14:13:29 The tracking symbol in this case is manipulated using a park on the graphics table. 14:13:39 In the two-handed version the targets were selected likewise, using the tablet. 14:13:44 However, navigation in this case was activated. using the left hand, using 2 touch sensitive strips. 14:13:53 The rightmost strip was used for smooth scrolling to the documents as demonstrated here, whereas the left strip was used to jump to the relative position in the document we just touched middle to the top of the doctor ones and to the bottom of the 14:14:02 document. From this view we see how the 2 hands are used together in performing the test class. 14:14:17 So there's a couple things that are subtle there that I think are really important. 14:14:22 Again we've actually note this paper was published the year after the Macintosh was released. The Macintosh had just come out. These These concepts were not known. 14:14:45 People Didn't know how to work mice very much even though they've been around for a couple of years. But let's put this way. 14:14:54 Microsoft sold a mouse in 1983 before Apple did that mouse in today' just for the mouse mit Ctl, and so very few people had mice because there was hardly any applications to use them but mice cost more than a 14:15:23 brand, and that's just for the for the mouse. So you can understand a few things about the economy of scale. 14:15:31 But second thing is mit Ctl and We notice this school bar that'll get more on this later has 2 strategies of how you navigate with the So we had 2 different strategies. The the user depending How far they had to go could jump up to the top and the bottom jump around 14:15:58 for in in the touch condition. and but so now let's look at the results that we got so experts. 14:16:08 Now this is a very There were some people who had used a tablet and a puck before very few people had. but there was a lot of people who had never seen a mouse or a puck before at that at this microsoft didn't have a successful width version of windows till 1,990 14:16:30 5, despite the fact that Macintosh came out in 84 mice were not universal. 14:16:40 Everyone's using dos and minorities so most users were were, and what people who worked with windows, but experts outperformed the ones who are beginners. 14:16:49 By 15% in in the whether it's the by manual versus the un manual technique novices in the two-handed group, i'll perform the one-headed group by 25%. 14:16:59 Okay So for the beginners. That meant if you're a complete novice, and never use a pointing device by using a bimanual technique, you actually performed faster, sooner than those who had the simpler soos called a one-handed technique and a lot of this can be be explained by time motion studies. because the time. 14:17:27 Moving between. But this is the one, and i'd forgotten this one to be honest until you know, when I was rereading the paper just to produce for this talk. 14:17:38 But I'm really happy to see this result, because what really is nice is that two-handed technique reduced the gap between the expert and novice. And remember, the whole thing I wanted to say is through chunking is you can aggregate 14:18:00 and make things simpler. So you just conceive of things as a single action rather than multiple things that you have to plan. 14:18:11 And there's all mit for the plausibility of my faith and the whole chunking theory and the aggregation. 14:18:28 And and now I want to just bring one other thing and to play here to show that the devices you use, even if all of this is true, and you have the same human beings for the same fingers doing the same task on the same machine one 14:18:43 100 at, you know, in the same day. I want to point out that this thing had these sliders, which is like this. 14:18:53 These bars here are like Mit, but a linear shaft encoder, whereas this thing is just a touch pad that we put a piece of paper. 14:19:15 It's really high tech We we really believe in Macgyver type of technology public duct tape and just whatever you can find kindergarten stuff put a piece of paper to Block I'm not gonna come towards the end of the book. here. So I can listen to the area. 14:20:34 So like a binary search and then I can and and that's what, and and that means that mit But that's exactly the kind of things that the scroll bars let you do and so you can't do that on 14:21:06 this because it's it's a motion sense of device mit ctl, and it can only do motion sensing it's like a mouse, whereas the digitizing tablet you can slide mit ctl and because the only the reason you can make this into a slider and that absolute is you just have to take 14:21:27 2. so any absolute device can become a relative device But a relative device can't necessarily become an absolute device. and and and so you there's a whole bunch of implications there. 14:21:43 But I just want to point out. every piece in the chain is important mit ctl, and there's a reason we and and most people would not different. 14:21:55 Take any book you take on input devices is probably not. 14:21:57 And by the way, Mit ctl and They're so different. And most people have never seen that type of Controller Brad, have you ever seen these use before? because I got them from Alice. And research. this is a Penny and Giles. 14:22:11 This is absolutely a really important device. It took me really long. 14:22:15 These are each one of these sliders. costs over a $1,000, because it's super high tech compared to these things we we got. 14:22:24 They're extremely plastic. but these are because you can actually move things and no matter what the parameter is if it's been adjusted by the computer, there's no handle there's a little light so you can just grab it really quickly without having to grab the handle and just nudge the value up or down to adjust 14:22:44 it. I have well over 800 devices like that in my collection. to illustrate this stuff, we have a catalog talk to Brown. 14:22:51 He'll play where he is so why not use 2 mice. This is this is the thing that drives me nuts. 14:22:59 So we said, Oh, yeah, 200. But we just use 2 mice. 14:23:02 That is almost almost always the wrong choice, We know that our hands are specialized for specialized tasks. 14:23:14 We've already seen the devices are specialized and have a It it's just got no personality. It's got no bias. 14:23:25 It's got no it's it's probably not and and so i'll give you some examples This is where what I'm talking about. 14:23:33 Now, mit Ctl and it's going to start showing why this stuff is still relevant today. 14:23:39 Why it's so important to know history the whole thing is about standing in the shoulders of giants there's a spreadsheet from today. I I did that snapshot today. 14:23:51 And okay, So what's it got there there's a scroll bar that scroll bar is identical to the one that Brad made for the Perk which mimicked the one that came on the Macintosh, the year 14:23:59 before he knew about them, because he'd been at xarks Park, and and so they weren't new to him. 14:24:06 But they were new to almost everybody else. So what does that think? 14:24:09 What do you call it? I just use the word the school bar one word while it's a hyphenated word. 14:24:16 No it's not it's actually 3 widgets let's stay constructed you have to know how to deconstruct it. assume everything is a hierarchy and and it's hiding to wrap up and the better wraps it up and the more invisible it is the 14:24:55 more First, there's the there's the elevator or whatever you want to call it, , and you can move that up and down, and it and it just and drop at any old place like Brad was dropping his fingers that rubs the top of the document to that position of the 14:25:14 document. Okay? Oh, there's smooth scroll you've got 2 controllers, one here to scroll down one way up there to school up. 14:25:22 Okay, Oh, there's a third one if you click in the middle you'll advance a unit typically one page up or one page down Mit, So there's actually 3 widgets there integrated into one hey? 14:25:54 Chunking, and you'd never how many of you ever thought of it that way, and actually could name those things. 14:25:59 And I explicitly identify the functions okay Oh, faith there's a hold account there's a whole other scroll bar at the bottom. 14:26:07 So. So now you've got the meta scroll bar with 3 functions. 14:26:15 So that's 4 tokens for 4 notes in the in the graph and the Bnf driver, and you've got another one over here. 14:26:21 So you've got 8 and just to describe the graph of that particular thing. If you're going to use a bn F. 14:26:27 Gram, or something conceptually. It should be one thing this works sometimes. but actually, if you were kind of big excel spreadsheet, which that is part of it is horrible, and nobody has started to think about it. 14:26:45 And here's how simple it is just from the experiments that comes directly the solution comes from stuff directly from what Brad and I showed that paper. 14:26:52 Here's a touch pad I made that you know restrontal. it's just that touchpad that you couldn't buy them hard at all. 14:27:00 1,994 Hardly anybody. the only people who had a touchpad with was a workstation company. who's Oh, man? 14:27:12 Came out of New England, and there were disaster at a touch pad, and it was mostly competent deployment, and it poisoned People's experience with touch pads. 14:27:29 Anyhow. There it is here's the functions that I just described smooth scroll jumped theater bowl, and go to a relative position. 14:27:39 The drag flick and tap you don't need multi-touch. 14:27:44 This is the other part not waste resources. some of the stuff you can do here. You can do in your apple pads, but use 2 fingers or one finger. They mean different things. 14:27:55 No, that's all moted anyhow. Why, would I waste multi-touch capabilities when I could do it with one finger. 14:28:00 So here's the here I can drag around and now, we'll just drag the document along that path up and down just like you were with the scroll. 14:28:10 Narrows smooth scroll. If I want to have orthogonal, purely horizontal, or purely vertical, like the frets on a guitar or the edge of a ruler, if I move my fingers along the edge i'm going to get pure horizontal and pure 14:28:26 ver mit. so you can go diagonally. All of a sudden you realize Holy Cow. I never thought of that. If you're using scroll bars and scroll arrows conventional ones. 14:29:07 You have to go up, then move over to the bottom page go a little bit left, and i'll get so I can go diagonal. 14:29:11 I can drag. So that's the first row smooth scrolling The jump interval is really simple. 14:29:14 Just go and flick and flick in the direction you want it to go, and it will go a unit. 14:29:20 Distance. click, click, flick up, down, diagonal. Whatever the game, simple, one gesture back, and the third is, we want to go to the top left corner. 14:29:28 You just touch and release and you can do all of that with your non dominant hand. and then your right hand it's still sitting there like napoleon's not down at hand. 14:29:37 So why wasted? Now put a pen in it, put a mouse in it, go to the touchscreen, but use it in concert with an appropriate thing that follows the priorities, and the skill sets to allocate the tasks according to the rules that guiard has given 14:29:56 us in such a simple way. so I have to go backwards again. 14:30:01 So I'm going back to 1,980 Okay, We're talking 5 years. 14:30:06 This paper was done there, and the work was done even earlier But this is like 7 years before the the the paper when this was happening. 14:30:30 What if 5, you know, brought it up? But this is a computer music instrument that I used to tour in fact, I've brought this down to Carnegie Mellon back in the eightys. 14:30:42 There were no graphics screens. There was an a calm 24 line Vk: 100 emulator. 14:30:49 The piano keyboard is a cleverly disguised Ascii device, Every key maps that. And so we just mapped. 14:30:53 Ask the end of the notes of the panel and could drive the synthesizer polyphonically in real time. 14:31:02 There you will see the slider box that we use for that. 14:31:07 Brad was using There's exactly the same tablet i've been using those things since 75 watch this compatibility Here we're using a straight Ascii African-american display but it's coupled with the outlook in such a way that we can make use 14:31:41 of I'll demonstrate this display has a moving cursor which tracks the position of the cursor on a tablet, so bottom up hand, corner, on the left hand corner. 14:31:59 Now information which is displayed on the terminal is in essence the operating console of the instrument. 14:32:02 There are a series of 8 scores, which information is set up in rows, and for each row I have a number of parameters which I can affect and real time. 14:32:08 Take score one and go to the particular control calls on up by pointing at the on-off switch and depressing the cursor button number one. 14:32:16 I can hear the score more plays once. and then switches out by moving to the next column for that score. 14:32:23 I'll turn the cycle, mode on which will result in the score, repeating once it's begun, We can now look at the other parameters, for example, working like the loud. 14:32:30 We have the grammar called richness. This is the richest of this background to that parameter. 14:32:35 I can entirely value. let's say the value 24 it's. 14:32:47 Now what we have done with would be increasingly risky. 14:32:49 However, I can use and change the value mean if you're adding in essence every number on the display is a grammar potentialometer exactly the same time as the graphic. 14:33:07 The university programs. That is my point, the value press. It does that a lot of people hold the value down as you can hear. 14:33:17 It gets clearer and push it out. okay compatibility here we're using. I want to just interrupt there for a minute. I should have pause. 14:33:29 The subway. Did I get it back? I want to to think about what you just saw. 14:33:37 In essence that console is a spreadsheet it's a it's rows and columns of numbers, and out Ascii characters, and anyone which is a scalar value I can point at and select just like you would any cell in a spreadsheet or any number in a 14:33:56 field, and then I can type it, but I can also, if I push and hold, drag its magnitude up and down. 14:34:06 And, furthermore, the further I go out when dragging, I get a very near effect. 14:34:11 So I control the resolution mit there's only one application that I've ever seen that actually, even though that's been published since 1,978 i've only seen that used once so i've got a full potentialometer as well, as a 14:34:49 selector, where I can get really high precision and accuracy with the keyboard. 14:35:00 But I can actually work and And explore things that's chunking and that and that's using analogs for the design straight Ascii African-american cursor button. 14:35:08 Number one for that score, for example, working to be increasingly risky typing. 14:35:11 However, mothers, and which I can use and change the value dragging in essence. 14:35:18 Every number on the display is a ground of potentiometer. 14:35:29 Exactly at the same time, was the grammar and geography programs that is like the value presses that a lot of the people and I can pull the value down as you can. 14:35:37 Hear it gets clearer, push it out functionally. We have neographic potentiometer, although the output rabbits are different, or the adaptive. 14:35:47 In fact, they only have an alphabetical display. do the same thing to amplitude, and quite an amplitude rang it out. Pull it down, or the articulation take a staccato very resonant, I could even change the 14:35:57 tempo, you know, which followed that little can, however, tighten up my wand. 14:36:11 Of course. No, and I have a button on the cursor which restores any parameter back to this default that is Bible. 14:36:23 2. So again, I get the yahoo minimum and i'll turn this off, and we'll point out another feature about the system We can control of those parameters independently for any one of a number of scores in fact, 8 scores can be playing 14:36:45 simultaneously, and each parameter can be independently controlled. Here we take the first voice. 14:36:47 We to which i'll have a second point now water partly I can control several parameters simultaneously by building up a hierarchical control structure that is i'll start both these scores together good 14:37:19 trigger connect them a single device. Then I can exactly now. 14:37:27 The 2 falls of the single ice spider drag them down beto other ideas that are built in the system, though in me to have the concept of an inverse group. 14:37:49 So I proved. One single gesture can increase the volume of one and decrease the other is a concept of phosphate in the express. 14:37:57 In a single gesture. Okay, so I just want to I'm. 14:38:11 Not trying to The only thing i'm trying to get across to you is where the ideas come from, how they evolve, how to get formalized, how you work from intuition how you work from history and how you Think the stuff through and how you see the relationship of the devices, the context and the 14:38:34 usage terms of gathering all of that, those data and starting to make intelligence strategies and techniques, to to advance and achieve something where you can make progress. 14:38:52 200. that was 1978, the that's the ipad The Macintosh came out in 84, and the Brad in my study, I think, was 85, or It was published in 8635 camera but it's 14:39:03 there. so this is why I already knew and I was trying to describe this stuff, but it didn't have the stuff with me. 14:39:21 Having done that. What that you know You could tap your head and then I just got rejection from people didn't believe me, People who really were smart. 14:39:28 I really respected, and I still respect them and so the thing here is notice it's the same slider box that Brad was using in the study. 14:39:38 It's the same tablet it's different tasks, but at one level of abstraction, or 2 levels of extraction, it's the same concept, it's all about chunking and be very erez agmoni that I can be moving the mouse doing one thing on one control and the physical controls at another because 14:40:06 the those are fixed position, and therefore through more memory. I can reach out and touch and grab them without looking just the way you can turn to radio while looking forward to the car because they're physically located. 14:40:21 They're not virtual things on a touchscreen but you have to look mit so for the decent past. 14:40:43 I'm going to take you to the future, because I want to wrap this up and say already there are things that in terms of scrolling and very simple things we do even with conventional laptops, we could make changes which are innovative and 14:40:59 could people at the same time I want. to point out that there's things that we can do, not just looking back, but looking forward where there are still high relevance in. 14:41:14 So there's a colleague of mine michelle Paiu. 14:41:16 He's showing an implementation of something called what he calls an in in hand. 14:41:25 Command. it's a palate. menu it's simply a tear off menu like you've seen on the Macintosh for years and on the But where you can hold it in your nondomer hand, and move it 14:41:34 around so the way to think about it is this my wife's a painter there's a painter's palate. 14:41:41 There's my stylus I want to make clear that i've been to the Reichs museum the museum of modern Art. I've been to museums all over the world art galleries and i'll tell you what I have never seen 14:42:08 a where they nailed the pallet up on the screen in order to pay right held. My wife is a painter. She did not pay this one. 14:42:19 I bought this one. Can you see any nail holes on it There's no nail holes. 14:42:24 Okay, Here is the painter's. menu this is the menu of colors. Here is the paint brush. 14:42:38 There is the canvas and guess what it should come to me then I move it out like that. I shouldn't have to go there or there, right Here's a hint going forward. 14:43:05 If there's any chrome on the edges of the screen widgets, the stuff that you have to go to you've already lost the game, because your interface design will have to change completely if because there's not enough pixels on the small displays and the pixels are too far away out of arms reach on the 14:43:23 big to minimize your applications and maximize accessibility and the relevance. The Ui has to come to you. 14:43:31 It has to be in hand. and disappear the minute this you're It's not that you can't use it. 14:43:40 You want to go back to dust? you want to go back to unix and have nothing on the screen except your work and have you work hover on top. So Michelle is taken a version of a palette. 14:44:00 Menu that's basically a tool glass sheet and he's touching onto a big screen. 14:44:08 We're getting the other side, and it just appears there, and I want you to notice that this is this notion about skill transfer is that if you use Powerpoint, you already know how to use that you know how to hold a hog to your hand on the 14:44:29 screen. You know how to have the menu appear with cup with your non dominant hand. 14:44:42 That menu. the iconography is exactly the same as what you've used for years there's nothing to learn except how bad the previous designs have been relative to what they could have been your job is to be frustrated with that not accept it and come up with a better 14:44:57 way, and then make it possible. So Here's a quick video to try and bring this forward. Hello! 14:45:30 I'm michelle Pierre and i'm going to talk about in place. Come on so in place. command. 14:46:00 Actually my news that i've got next to your fing So you know the second menu is actually a 3 last menu. 14:46:07 So it's like a key to so I can basically, if I want to go Align, I can just go with my friend here. stuff. Might I? 14:46:16 Yeah, I know I can position my line with 2 hands on the finger side. 14:46:22 I have an upset. so my CEO doesn't high at the beginning of the line. 14:46:30 The other side is just a band that shows the the positioning of the line, and then, when I release the menu will come back. And now I can do the same thing. 14:46:41 What else? if I want to do a square and start by taking 2 square, and I can position and size the square of this impact same idea for us, position and size, this Fpl. 14:47:00 We do have, which allows me to position it very precisely. well useful there. Because yeah, well, every day to get something to stop. Okay. time to stop. 14:47:07 Okay, i'm finished. I want to discuss we're all using very expensive computers. 14:47:14 Things are relevant there. This is a elastic band. 14:47:17 I just went to get it. Okay, when you use everybody's use elastic band before right? 14:47:29 And if you go to your computer what they call a rubber band line isn't to rubber band, it's a catapult because one end is anchored and you cannot move it at the same time. here. 14:47:46 I can shape it in different shapes move it around it's malleable. 14:47:49 But you need 2 hands, and what you saw there it is. He comes sooner. 14:47:53 The tool palette. he clicks through sweeps out the top right hand corner, for example, because he's already in the middle of something. 14:48:05 The menu disappears, and now his left hand is attached to the bottom left hand corner, so now he can stretch it, change its aspect, then release it. 14:48:09 One single connected chunked gesture Mit, and it just go matter on that. 14:48:21 So here we go i'm sorry went too, long the I just want to say there's history behind there's history going forward, and there's the stuff where we are right now. 14:48:32 You're highly positioned, if you take advantage of what's already out there to make a huge difference, and I hope from the examples. 14:48:42 The approach the methodologies, but also the thought processes are the principles that I've talked about today. 14:48:44 We'll give you a lake up so that when I retire because i'm 73. 14:48:48 I'm not going to be around for that much longer. 14:48:50 That you can make a better world for my grandchildren. and and to actually bring the stuff forward because there's so much need. 14:48:56 Right now you want to shrink the digital, divide, the economic divide and the educational divide. 14:49:01 This is to me the only way, you can do it mit ctl and find the most economical way to learn the most economical way to deliver the technologies and make sure you' help the people who are most in need in so doing as we know in universal design you'll also help yourselves and the people who 14:49:14 who don't need it's badly because you'll get the Connors of scale. 14:49:19 So, thank you, and that's really for your attention. and I hope this helped, and I will take questions by email if you choose to do that, because I remiss in not having given you a chance. 14:49:29 Well, I think we'll let people go if anybody wants us to put orhand so I think we'll let the students go. 14:49:42 I wanted to ask you about apples transition to school bars without the indicators on the top and bottom, and without the ability to drag the indicator as A Microsoft Employee. 14:49:57 Now you have the opportunity to tell us what you think about that transition. Okay. 14:50:03 So. so I have I. I have multiple answers one. I am not a Mac dush user, and I do so. 14:50:17 I, my most. my family are so I will have them. 14:50:24 Show me that wife is just replacing her mac because she can't run the new system. so it's actually in progress this very minute. if I did know I probably wouldn't express because I do not like to comment on negatively if if I did have a negative reaction on my competitors. 14:50:52 I respect them too much, because until my company's house is in perfect order. It just becomes a slag and faster. Who wants to do that? 14:50:59 We we have to it's like okay it's the Olymp the limit, sir. 14:51:01 Why would you want to win the grand slum, or something, or this space skating? 14:51:06 I guess people who are incompetent if i'm as a competitive athlete I have competed internationally on a couple of sports. I want to compete against the best. 14:51:17 Because That's How I measure that's where my ambitions are. and so I have huge respect for apple and and and lot of the other companies. 14:51:24 There's lots of things I don't like and There's a lot of things I don't respect but there's no stupid people there, and the question is how do we create a culture, and the people that are working there might be 14:51:44 working in my company at some of the time it's just we have to look at this collectively. 14:51:50 Otherwise we just get the 200. The partisanship part of my belief in that. Brad will tell you this. 14:51:54 There's not a single thing that I have ever done or people in my research group have done at the university, which has ever been patented, because I kick you out of my group. 14:52:04 If you patent because it's a public service public pace with universities and and by the way i'd rather spend the money. 14:52:16 It would cost a patent on supporting another graduate student or 2. That's just my value system. i'm not trying to impose it. 14:52:22 But that's just how I work. So the thing is is that I view this as a community. 14:52:27 And and I don't care where you go if you go to Apple or or Facebook or Google or Amazon, or or any startup. 14:52:38 And you take these ideas, and you make them huge 2! 14:52:39 God bless you! I believe in the ideas of. And so the final thing about your question is, 14:52:46 The more you know about how large companies work, and what the vested costs are, and so on. 14:52:54 And and you start to actually appreciate that whereas we can revise code we make. 14:53:01 We might, if I said, anybody here worked on a team, that's look greater than 10 people to build an application or anything. 200 try try working on a team of 4,000 people. I think you find it people working on office for example, 14:53:18 probably just just engineers are probably around 3,000 I don't know the exact number, but it could be more than that. the amount of code there that you have to support make sure it's perfect. 14:53:26 So that you have a 1,000,000,000 people a day using your code? 14:53:33 You don't want to take chances this notion of here. let's Let's just build it try it. 14:53:37 If it doesn't try that you don't want your person managing your bank account to use that code. 14:53:46 And so there's these different things that actually change the dynamics, and that's why i'm frustrated. 14:53:53 But I but I balance that with a certain amount of understanding. 14:53:59 And then the question is, how do you become effective? in engineering, how your engineering style ideas get engineered type of thing. 14:54:10 You know It's about social engineering, economic engineering, and and those are the parts that go into. 14:54:13 How how we move ahead and and it's just but that does The reason i'm still working Brad is because I don't need to, is because I still care the stuff's. 14:54:25 Important enough that I can to afford not to given that i've got 4 grandchildren and and and also I love the people I work with, and I think they're well intentioned. 14:54:38 It's just really hard and I think these really hard problems take really serious people. 14:54:41 But i'd also say the most you cannot be creative without play. 14:54:53 And so the most important thing is is that how do you take a relaxed, playful attitude. Oscar, i'll finish with this, Oscar Wilde said. 14:55:00 This great thing. These issues are far too important to Take seriously and because you can't be creative when you're just scared to death.