UC DLFx 2018: Keynote by Don Norman
UC DLFx 2018: Keynote by Don Norman
Broad experience.
"The best way to predict the future is to invent it." ~ originally from Nobel Prizewinner, book published in the 1940s. Imposible - no matter what we invent it never turns out the way we want it to. Easy to predict future, hard part is getting it right.
What he does know about libraries. Rainbow's End, The Name of the Rose. Engelbart's Groundbreaking demonstration of what you can do. Read all sorts of reports, asked lists. Nothing he can tell us we haven't already heard, thought, or do. He should say what he does and how he works and what it is as a designer and what he needs. Design is a way of thinking, of approaching problems. Trying t teach this to all - democratizing design. Don't need to send in experts to tell you your needs and build something to solve needs, and convince you this is what you need. This is standard. Instead, show you how we learn and let you do it yourself with mentoring and tutoring.
Libraries up to now are mostly a place where people go to learn the knowledge of the past. Let's make it so libraries are a place where people create the knowledge of the future as a way people ca interact with it and experiment with it. Nowhere else in the university. Makerspace is for physical products. Why doesn't library become mentors who know what knowledge is, where, and how it can be applied. You should be able to move graphs. Consumer Reports - we can't change importance of weights of actors, but you can't touch their data, even on their website to be able to view it as you like. We shouldn't just have a PDF or snapshot in repository that loses dynamic ability. Libraries: keep historical data but make it dynamic. Data is continually modified after publication - what if someone wants the early version? University researchers are good at what they do but they don't do that much. In the university, tremendous amount of great deep thought but little action .In industry, a tremendous amount of action, but no thinking. Pace of industry is different so no time for reflection on product put out because need to put out next thing. At university, what if we organized by problems instead of disciplines. It would allow people to work interdisciplinarily. As the Library most of us are practitioners who get stuff done, and library field is one of practice. Place where people can learn to put it together.
A friend teaches a course on search. He has a blog - Dan Russel. Gives problems that never occurred to him that you can search for an answer. Gives answer the next week. Library should be the place where new exciting things happen. City library has a different mission, and he hasn't seen an academic library that excites him. Turning libraries into study halls.
Elevator and a train - what in common.
Why scolded for using a computer terminal back in the 60s to write his paper. Why did Apple and Microsoft miss the coming of the internet?
Human centered design. Understanding the world vs creating making things beautiful. But who designs our info systems, or if first job is to redesign health system .Where does design fit? What we really want is objects that are beautiful but we also want experience in using to be enjoyable and pleasant. Behind scenes everything needs to work properly. Design is a practice and way of thinking, and is across disciplines. Firs problem is focus on people who are involved. At nexus between people and technology. Technology is anything created by humans for use by humans. Paper, pencils, etc. New technology tends to be difficult to use and confusing, from 1600 - difficult to use modern plow. Second thing is to solve the right problem. Whatever problem they ask me to solve, I do not solve. What is underlying issue?
I'm not going to tell you about digital libraries and digitization, you all know that. We insisted on doing systems approach-everything is interconnected. Local optimization does not lead to global optimization. Small fixes can make things work. Even after deep thought ad present solution, probably wrong, so quickly prototype (hours) and test quickly. First PDA was cut to wood just to see if size was convenient.
Students walking on campus just cross street, meanwhile, lots of cars. During class breaks, car cant get through mob. e need to give more intention than blinkers if you want to use autonomous vehicles. can't be trusted. Went and bought upholstery - driver puts on 'seat suit', person can see, from outside can't see driver--got IRB approval, haha. Now widely deployed - originally at Stanford.
Train stops at every stop, elevator doesn't, so maybe elevator is more like a bus (horizontal vs vertical). Now inefficient - destination control elevator: will tell you which elevator to wait for, a supposed to random people getting on and stopping at every floor. But people hate it--waiting at elevator F but G keeps coming and going. People hate it. No controls if select wrong...just get out when it stops. Make a change, is hard for people to get used to.
Why do you catalog? So people can find them. Spend a lot of time categorizing them. Amazon used to put same with same like we do. Now they put it wherever it fits. Why do we alphabetize things? If digital, not necessary, just ask for what you want from internet. Catalog - put wherever it fits and wherever it's easiest. To find, now go to system with classification code, then what floor of library. Tell them 8th floor here, but related book may be elsewhere. One of the things we have to do to change the world is rethink the world. If a book is misplaced, basically lost it. New scheme still has that issue. But pluses and minuses, most important is to rethink what the issues are, and old way nutncessarily best way. MS and Apple missed internet because they are too expert. To be a leader in the field: be the only one. But can't be only field no one cares about, has to be exciting new area. New directions for design, what design could be and applying it. Then when everyone else says yes in the world, then he leaves.
He was at Apple, MS missed internet. When first UIUC developed browser - www was just being invented at CERN but hard to use. Students decided to invent something usable. When came to Apple to advanced technology group and said hey, we have interesting browser thingy you paid for, why not make into a product. Apple said no, you need two way pointers, and you guys only have one way pointers. So started own company and bought out Netscape. Apple was right, you do get lost on the internet, but so what? You don't have to have correct answer, have to have something that makes a difference--knowing too much can get in the way. Miss new things because so expert in the old way.
Scolded for typing paper in 60s - precursor of ARPANet. Wasteful - using expensive computer to type paper, also each character has to be sent away to be determined to be correct. Desire lines: the paths people take off the sidewalk. Architects put up fences, lakes, fences to avoid desire lines. People using library to study is a waste of resources. But there is a real need for that, so we should be rethinking what we are doing. Sol study, group rooms, learning centers. Maybe not best use of library space and librarians but is a need for the campus. When people doing things you don't think they should do should take as a signal to move forward.
What should the digital library be? If you take a look at the way we work--what is the real problem here, and how can ewe make it better and more efficient for the people (not machines administrators). Why do we do 3 hours class each week 10 weeks? Not how learning works. Learning comes in packets. But easier for registrar to assign rooms. Lecturing is easier for professor to get through material, but not a good way to learn. We do it because of the system. We build a thing, doesn't work completely, we fill in the missing pieces so human labor replicates what machine should be doing. Driving is a good example of something that shouldn't be done by people - hard because took a month or two to learn to drive and a year to be comfortable; simple but have to respond accurately and precisely. Take 10-20 seconds to figure out what to do that's so many feet before response. Flying simulator - the easy ones were difficult, the tough ones were so computerized but pilots take 3-5 minutes of what to do, but really well trained. Drivers not as well trained. Why do we ask people to act like machines and work by machine rules. We should design for people abilities and design for things people are bad at. That's one issue.
But we also want to let other people figure out stuff. Like Wikipedia - issues is that you can only use secondary sources and no primary. But still a positive source of information. at Encyclopedia Brittanica, experts wrote papers - getting to agree, then the writing, then fact checking, then edit to make accessible to readers. But should let people write in first person, let them write their story, let them write their strong opinions (and encourage others to write rebuttals). But most people go elsewhere for the truth, not encyclopedia. Amazing how little truth there is. Countries are still in dispute over boundaries. Is Taiwan its own country or a district of China? No firm rules.
Democratizing knowledge and design. Wikipedia allows different ideas from different people so long s they cite. But wikipedia has its own fashion of knowledge as if it can be write-in text.
Distilled - AI journal out of Google. All the diagrams work and as you ready can plug in numbers and see how changes, or delete feedback loop to see how it changes. Problem with the journal is it takes a year to write an article because diagrams are hard. So we need new tools to make it easy to do and then easy to publish in a sensible way. This is role of library. Visiting library...not the building. Science and humanities scalars are finding new ways of publishing. He publishes more in LinkedIn because doesn't know where to publish and he gets huge audience there. We are the ones who have to do and lead and mentor the changing.
Library is a place where used to examine knowledge of the past, but should be place to create and disseminate new ways of developing knowledge for the future.
Questions
Q: Designing for humans, but we design for ourselves and not end users. We design the standard for sharing geographic data in MARC with thousands of elements but only use a few. How can we retrain our empathy to look at users instead of selves as community we serve?
A: Gets to the heart of many professions. Medical profession - specialist knows so much, they don't see you as a person but symptom. Just one symptom - other specialists will deal with other symptoms. Engineers cant understand why people don't like the design. Hard when you know too much to imagine. Lots of observation, no questionnaires or surveys. (i.e., not patient focused help which frustrates physicians and nurses and technical folks - that's where issues are, make the staff happy). Science museums - if you read tags on items, not interesting things, but where the got it, what year, who donated it--written for other curators and not the public. Not that you shouldn't write for curators, but should ALSO do for public. Error messages on computers - shouldn't say error, should say what it is and what to do to fix it, but when you do, instruction disappears so you don't remember....as much as possible should allow person to do. Thinking about how people use things, the questions they ask--what answers do they really need? Engineers and folks should listen to the helplines, and don't assume they're stupid, think of it as something to solve.
Q: Democratizing Design - how do you so this without design by committee.
A: In Portugal, give 3D printer to kids and let them design and fool with it. Not done by committee. Group of people in Scandinavia - participatory design - design by committee creates issues. Rather, giving you the tools to do your own stuff. People with diabetes have to always monitor blood sugar level, a pain to prick fingers and determine what you should eat, how much insulin. Now a continuous glucose monitor. Another company makes an insulin pump. But why can't see on cell phone? Someone reverse engineered and published API on internet --easy to monitor their own, to track children. Why not connect the two? Soon developed an artificial pancreas. A little bit of committee like Linux - a few folks t top to make sure it's robust. Allowed to build it for themselves to make life easier for them. Companies can't do it - FDA approval is crazy expensive and long term, and trials have folks drop out. But people did and companies encouraged them; companies got together and released artificial pancreas, got through FDA more quickly. Design by committee is always compromise. With today's tech we don't need to compromise because we don't need one design to fit a million people. If we decrease cost, is possible to have a million designs. See IKEA.
Broad experience.
"The best way to predict the future is to invent it." ~ originally from Nobel Prizewinner, book published in the 1940s. Imposible - no matter what we invent it never turns out the way we want it to. Easy to predict future, hard part is getting it right.
What he does know about libraries. Rainbow's End, The Name of the Rose. Engelbart's Groundbreaking demonstration of what you can do. Read all sorts of reports, asked lists. Nothing he can tell us we haven't already heard, thought, or do. He should say what he does and how he works and what it is as a designer and what he needs. Design is a way of thinking, of approaching problems. Trying t teach this to all - democratizing design. Don't need to send in experts to tell you your needs and build something to solve needs, and convince you this is what you need. This is standard. Instead, show you how we learn and let you do it yourself with mentoring and tutoring.
Libraries up to now are mostly a place where people go to learn the knowledge of the past. Let's make it so libraries are a place where people create the knowledge of the future as a way people ca interact with it and experiment with it. Nowhere else in the university. Makerspace is for physical products. Why doesn't library become mentors who know what knowledge is, where, and how it can be applied. You should be able to move graphs. Consumer Reports - we can't change importance of weights of actors, but you can't touch their data, even on their website to be able to view it as you like. We shouldn't just have a PDF or snapshot in repository that loses dynamic ability. Libraries: keep historical data but make it dynamic. Data is continually modified after publication - what if someone wants the early version? University researchers are good at what they do but they don't do that much. In the university, tremendous amount of great deep thought but little action .In industry, a tremendous amount of action, but no thinking. Pace of industry is different so no time for reflection on product put out because need to put out next thing. At university, what if we organized by problems instead of disciplines. It would allow people to work interdisciplinarily. As the Library most of us are practitioners who get stuff done, and library field is one of practice. Place where people can learn to put it together.
A friend teaches a course on search. He has a blog - Dan Russel. Gives problems that never occurred to him that you can search for an answer. Gives answer the next week. Library should be the place where new exciting things happen. City library has a different mission, and he hasn't seen an academic library that excites him. Turning libraries into study halls.
Elevator and a train - what in common.
Why scolded for using a computer terminal back in the 60s to write his paper. Why did Apple and Microsoft miss the coming of the internet?
Human centered design. Understanding the world vs creating making things beautiful. But who designs our info systems, or if first job is to redesign health system .Where does design fit? What we really want is objects that are beautiful but we also want experience in using to be enjoyable and pleasant. Behind scenes everything needs to work properly. Design is a practice and way of thinking, and is across disciplines. Firs problem is focus on people who are involved. At nexus between people and technology. Technology is anything created by humans for use by humans. Paper, pencils, etc. New technology tends to be difficult to use and confusing, from 1600 - difficult to use modern plow. Second thing is to solve the right problem. Whatever problem they ask me to solve, I do not solve. What is underlying issue?
I'm not going to tell you about digital libraries and digitization, you all know that. We insisted on doing systems approach-everything is interconnected. Local optimization does not lead to global optimization. Small fixes can make things work. Even after deep thought ad present solution, probably wrong, so quickly prototype (hours) and test quickly. First PDA was cut to wood just to see if size was convenient.
Students walking on campus just cross street, meanwhile, lots of cars. During class breaks, car cant get through mob. e need to give more intention than blinkers if you want to use autonomous vehicles. can't be trusted. Went and bought upholstery - driver puts on 'seat suit', person can see, from outside can't see driver--got IRB approval, haha. Now widely deployed - originally at Stanford.
Train stops at every stop, elevator doesn't, so maybe elevator is more like a bus (horizontal vs vertical). Now inefficient - destination control elevator: will tell you which elevator to wait for, a supposed to random people getting on and stopping at every floor. But people hate it--waiting at elevator F but G keeps coming and going. People hate it. No controls if select wrong...just get out when it stops. Make a change, is hard for people to get used to.
Why do you catalog? So people can find them. Spend a lot of time categorizing them. Amazon used to put same with same like we do. Now they put it wherever it fits. Why do we alphabetize things? If digital, not necessary, just ask for what you want from internet. Catalog - put wherever it fits and wherever it's easiest. To find, now go to system with classification code, then what floor of library. Tell them 8th floor here, but related book may be elsewhere. One of the things we have to do to change the world is rethink the world. If a book is misplaced, basically lost it. New scheme still has that issue. But pluses and minuses, most important is to rethink what the issues are, and old way nutncessarily best way. MS and Apple missed internet because they are too expert. To be a leader in the field: be the only one. But can't be only field no one cares about, has to be exciting new area. New directions for design, what design could be and applying it. Then when everyone else says yes in the world, then he leaves.
He was at Apple, MS missed internet. When first UIUC developed browser - www was just being invented at CERN but hard to use. Students decided to invent something usable. When came to Apple to advanced technology group and said hey, we have interesting browser thingy you paid for, why not make into a product. Apple said no, you need two way pointers, and you guys only have one way pointers. So started own company and bought out Netscape. Apple was right, you do get lost on the internet, but so what? You don't have to have correct answer, have to have something that makes a difference--knowing too much can get in the way. Miss new things because so expert in the old way.
Scolded for typing paper in 60s - precursor of ARPANet. Wasteful - using expensive computer to type paper, also each character has to be sent away to be determined to be correct. Desire lines: the paths people take off the sidewalk. Architects put up fences, lakes, fences to avoid desire lines. People using library to study is a waste of resources. But there is a real need for that, so we should be rethinking what we are doing. Sol study, group rooms, learning centers. Maybe not best use of library space and librarians but is a need for the campus. When people doing things you don't think they should do should take as a signal to move forward.
What should the digital library be? If you take a look at the way we work--what is the real problem here, and how can ewe make it better and more efficient for the people (not machines administrators). Why do we do 3 hours class each week 10 weeks? Not how learning works. Learning comes in packets. But easier for registrar to assign rooms. Lecturing is easier for professor to get through material, but not a good way to learn. We do it because of the system. We build a thing, doesn't work completely, we fill in the missing pieces so human labor replicates what machine should be doing. Driving is a good example of something that shouldn't be done by people - hard because took a month or two to learn to drive and a year to be comfortable; simple but have to respond accurately and precisely. Take 10-20 seconds to figure out what to do that's so many feet before response. Flying simulator - the easy ones were difficult, the tough ones were so computerized but pilots take 3-5 minutes of what to do, but really well trained. Drivers not as well trained. Why do we ask people to act like machines and work by machine rules. We should design for people abilities and design for things people are bad at. That's one issue.
But we also want to let other people figure out stuff. Like Wikipedia - issues is that you can only use secondary sources and no primary. But still a positive source of information. at Encyclopedia Brittanica, experts wrote papers - getting to agree, then the writing, then fact checking, then edit to make accessible to readers. But should let people write in first person, let them write their story, let them write their strong opinions (and encourage others to write rebuttals). But most people go elsewhere for the truth, not encyclopedia. Amazing how little truth there is. Countries are still in dispute over boundaries. Is Taiwan its own country or a district of China? No firm rules.
Democratizing knowledge and design. Wikipedia allows different ideas from different people so long s they cite. But wikipedia has its own fashion of knowledge as if it can be write-in text.
Distilled - AI journal out of Google. All the diagrams work and as you ready can plug in numbers and see how changes, or delete feedback loop to see how it changes. Problem with the journal is it takes a year to write an article because diagrams are hard. So we need new tools to make it easy to do and then easy to publish in a sensible way. This is role of library. Visiting library...not the building. Science and humanities scalars are finding new ways of publishing. He publishes more in LinkedIn because doesn't know where to publish and he gets huge audience there. We are the ones who have to do and lead and mentor the changing.
Library is a place where used to examine knowledge of the past, but should be place to create and disseminate new ways of developing knowledge for the future.
Questions
Q: Designing for humans, but we design for ourselves and not end users. We design the standard for sharing geographic data in MARC with thousands of elements but only use a few. How can we retrain our empathy to look at users instead of selves as community we serve?
A: Gets to the heart of many professions. Medical profession - specialist knows so much, they don't see you as a person but symptom. Just one symptom - other specialists will deal with other symptoms. Engineers cant understand why people don't like the design. Hard when you know too much to imagine. Lots of observation, no questionnaires or surveys. (i.e., not patient focused help which frustrates physicians and nurses and technical folks - that's where issues are, make the staff happy). Science museums - if you read tags on items, not interesting things, but where the got it, what year, who donated it--written for other curators and not the public. Not that you shouldn't write for curators, but should ALSO do for public. Error messages on computers - shouldn't say error, should say what it is and what to do to fix it, but when you do, instruction disappears so you don't remember....as much as possible should allow person to do. Thinking about how people use things, the questions they ask--what answers do they really need? Engineers and folks should listen to the helplines, and don't assume they're stupid, think of it as something to solve.
Q: Democratizing Design - how do you so this without design by committee.
A: In Portugal, give 3D printer to kids and let them design and fool with it. Not done by committee. Group of people in Scandinavia - participatory design - design by committee creates issues. Rather, giving you the tools to do your own stuff. People with diabetes have to always monitor blood sugar level, a pain to prick fingers and determine what you should eat, how much insulin. Now a continuous glucose monitor. Another company makes an insulin pump. But why can't see on cell phone? Someone reverse engineered and published API on internet --easy to monitor their own, to track children. Why not connect the two? Soon developed an artificial pancreas. A little bit of committee like Linux - a few folks t top to make sure it's robust. Allowed to build it for themselves to make life easier for them. Companies can't do it - FDA approval is crazy expensive and long term, and trials have folks drop out. But people did and companies encouraged them; companies got together and released artificial pancreas, got through FDA more quickly. Design by committee is always compromise. With today's tech we don't need to compromise because we don't need one design to fit a million people. If we decrease cost, is possible to have a million designs. See IKEA.
Comments