Duration 35:42
16+
Play
Video

Design, machine learning, and creativity

Douglas Eck
Principal Scientist at Google
+ 3 speakers
  • Video
  • Table of contents
  • Video
2018 Google I/O
May 8, 2018, Mountain View, USA
2018 Google I/O
Video
Design, machine learning, and creativity
Available
In cart
Free
Free
Free
Free
Free
Free
Add to favorites
8.1 K
I like 0
I dislike 0
Available
In cart
Free
Free
Free
Free
Free
Free
  • Description
  • Transcript
  • Discussion

About speakers

Douglas Eck
Principal Scientist at Google
Brenda Fogg
Head of production at Google
Ryan Germick
Principal designer at Google
Isabelle Olsson
Design director at Google

Doug leads Magenta, a Google Brain project working to generate music, video, image and text using deep learning and reinforcement learning. A main goal of Magenta is to better understanding how AI can enable artists and musicians to express themselves in innovative new ways. Before Magenta, Doug led the Google Play Music search and recommendation team. From 2003 to 2010 Doug was faculty at the University of Montreal's MILA machine learning lab, where he worked on expressive music performance and automatic t

View the profile

Brenda is head of production at Google Creative Lab, an interdisciplinary team of designers, technologists, filmmakers, writers, strategists, and more. The group experiments with innovative product and marketing projects, often collaborating with other teams across the company on new ways to connect people to Google's products. She got her start in technology at Microsoft in the late '90s, then spent ten years in creative agencies telling stories and creating brand experiences prior to joining Google. Brend

View the profile

Ryan Germick serves as the principal designer of Google Doodles and the Google Assistant’s personality. He and his team of artists, writers, and producers aim to make technology more accessible through creativity and craft.

View the profile

Isabelle Olsson is the design director for the Google Home and Google Wearables portfolio. She also oversees CMF (Color, Material and Finish) for all of Google hardware. Isabelle joined Google in 2011 working inside inside Google X, turning new technology into compelling prototypes. Prior to joining Google, Isabelle worked at Fuseproject designing a wide variety of products, ranging from consumer electronics, exhibitions, packaging and jewelry. She has also designed several furniture pieces exhibited at Mil

View the profile

About the talk


Share

I'll tell people that work at cool designs. What is there to design? Cleveland The Judds Oh, I know it's cat. Hi, how can I help? Hello everybody. I'm friend of fog. I work in the creative lab which is a sort of a multidisciplinary Creative Group within Google that often collaborates with other teams across the company on products or technology experiments and sometimes finding new ways to talk about some of the Innovative work that's going on inside of Google and at some point over the years. I've

worked with each of our panelists on some of those projects that you probably saw on the video just now, so we're going to talk a little bit about some things that you may have seen a little bit about design at Google and what that means is a sort of a fundamental Framework and connective tissue between the things that Google makes and making them as useful and accessible to as many people as possible start with some introductions. We have dug act right here to my left who leads a project called magenta

which is all about exploring machine learning and creativity with Isabelle Olsson who is responsible for the design of Google home and wearables and hardware and over there is Ryan germick principal designer who also is known as the doodle guy and has been quite involved in the Google assistant and generally just in the business of the lighting users everywhere. So let's let's start by letting everybody talk a little bit about what you do. I could almost start with dad cuz you're sitting next to me you head up this project called magenta for anybody who doesn't

know exactly. What side is me? If you can talk a little bit about that. It may be touched on what inspired the the group in the focus of of the team or crasher. Free music in our generation and we started by trying to understand the capacity to use a I specifically machine learning to generate music and art and it took us about a month to realize that that's asking the wrong question because if all you're doing is trying to generate music and art then you just keep pushing

this button and the Machine learning keeps making music and art for you and you know, it gets boring fast. So we've hit a very quickly to talk about how can we use machine learning to enable artists and enabling musicians to make something new to make something different. So it's it's sitting on the same idea of technology in art interacting with one another in a virtuous way starting, you know, with cave drawings and moving forward through the film camera and other bits of technology and so you're not necessarily trying to

replace creativity or duplicate creativity, but more providing the tools to enable people to do that. I think it's not just because that's what we choose to focus on. I think creativity is fundamentally human and it's about communication. So if we take the communication Loop out we can imagine in some analytic away a computer generating new things. But what makes creativity work is how we respond to it and how we then feed back into that process. So I think it's very much a society Society call a communicative Act of creating new things like

maybe some things that weren't necessarily possible before but like burnt humanly possible to create so there's nothing to talk about this with this example of that was shown in the in the lead up into the keno yesterday the incident. Project which is one of those things that's conservative arguments what human creativity can do want to touch on that sofa and Santana the following Hardware devices Ensign super you may have seen was played on stage before the keno to discuss their I might think the main idea

there is can we use machine learning to allow us to generate new sound and sounds on musically meaningful to us. And one thing to point out is that you know, we already have ways to do that, right? There's a bunch of great software. We also like I have a piano at my house so I can have a guitar they lots of ways to make some what we hope we can get it some kind of expressive edge with a eye something that we can do with these models a kind of intuitiveness or a kind of a new kind of Mobility artistically if I have a new tool and the one thing I would say, I don't want to take up too much

time cuz you know, there's a lot of other great artistic. If I said it was treated as something to capture reality and it was transformed into an artistic device by artist by photographers and I hope I'm a dentist that we find the right artist and the right musicians to take what we're doing and turn it into something creative creative as a bell when you're designing a product the home for example Year trying to create something that appeals to everyone through its design but

everybody's different right. So there's you know, these are physical products that share physical space with the people that use them in like sum in sometimes you have to cohabitate to talk a little bit about how you approach that problem. Yeah, I mean, I think I have the utmost respect for for people's homes. And like you said they're they're all difference and I think next to your body. Your home is your most intimate space and is the place for you share with your loved ones and your family. So to enter that space with with our products, we have to be super thoughtful about what we do

there and I think for us the most important thing is to be inspired by the context in which are products live in so when we were designing Google home mini, for example, the goal was to design an assistant for every room and that means your bedside table and your bedside table. That's where you put devices that help you see better or like a book that helps you dream. So that space is just so special. We wanted to create something that that was beautiful that's fit into the home. And

didn't take up too much attention and kind of faded in a little bit into the background and you're also responsible for CMS at Google which is Keller material finished and I've heard this ever heard the story about testing like a hundred and fifty different versions color palettes for the mini is that is that right for us the developing the color palette for for the Google family of products and individual products is a very It's a combination of Art and Science I would say and we

start usually 2 to 3 years before the products come out. So we have to do a lot of anticipation of Rights societies going right friends are going and take all of those kind of inputs into account how to make sure that when we release a product if it makes sense to people in addition to that, of course. When you design for the home, you have to think about the fact that you know, there's going to be light hitting the product. How does it stand the test of time we want to make sure the product beautiful for you for long time. So we have to go through a lot of iteration to get it right

and then also especially as we're developing fabrics, for example, depending on where you put it it takes different. It looks different in different lighting conditions. So when we ruin the redesign mini we went through I think hundred and fifty iterations of just the gray color. So but it was a lot of fun and it was about finding that right balance with what is too light. What's too dark and the other day I got this lovely email by I'm someone on the team with picked out his couch to match Google

home Max. So I took that as a giant compliment because we were trying to do the other way around But that was that was a beautiful Story how much intersection of the intuition that you use as a designer when your approach these kinds of problems with the sort of iterative testing and and sort of the scientific materials examination usually but we have fun with it. I think I think the key is gather as much input as possible and then digested and then come up with prototypes and way of ways of relating to

how this will fit into people's home. So even right next to my desk, I have a big bookshelf that we plays random object from all over the world for inspiration, but also to kind of put our stuff there quickly to see how does it feel and how does it feel overtime because it's not only about creating something that you are first attracted to but it's has to be thing. That you can live with for a long time. The Ryan you leave the Google doodle team and team is unique in a lot of ways namely one of them is that you

regularly and willfully break the brand rules for four times and that's unusual because it's you know, what the core of the brand and that's something that that seems to keep working and working working over the years. So talk a little bit about why that's why you think it's important to have the ability to kind of mess with it. I'm sure of his mission is to organize the world's information and make it universally accessible and useful. And I believe in that mission. I think it's a very powerful good thing to do for the world. And we hone in on the idea of

making things accessible by creating an emotional connection with users and sometimes like mucking up the standards if that's like lateral damage for people getting a positive charge and and learning something new or having fun with something that we think it's worthwhile. And yeah, I think human love all those things that are more important to you know, then it doesn't see and it's R Us more about using our creativity and crafts to make people feel welcome to the space at technology. So making people feel welcome in this beso technology. You also lead the team who created the

the personality for the Google assistant and Like how do you create a personality of the the the transactional things that have to happen between a user when they're interacting with an assistant digital assistant and they have an expectation that they're going to be delivered the information that they asked for and you felt like I needed to go a little bit farther than that to the transactional relationship but people have you know little bits the way we were talking with Isabella. Do you know everyone has sort of different things that they

like to interact with and some people like small talk and some people don't and some people think things are funny that other people think are totally not funny at all. So how did you talk a little about that technology like the assistant that comes in a smart speaker or smart display or in your phone is really personal that's one thing and so we just recognized that we have a different set of design challenges and if it was like more like objective like a Google search engine might be seen and then also You know when you would like this technology and share life

with you know, we're using this conversational interface as a metaphor. You can talk to it and I can respond to you in as soon as you hear the human voice opportunity to have a director, but you know, but we seen it so much like an obligation to design for the character cuz if you don't design for the people will just assume you're you don't you don't have much of a character but there's still some in place of character that we had from noodles and being, you know, I didn't post it character for Google where we celebrate certain things and we get creative and nerdy and excited and we tried a

chance that over to the Google Assistant where could be like a character that you'd want to spend time with because it has you know things have to get excited about or its you have the perspective of really wants to help you and not just be something that you want to use with something that you want to spend time with. So so surprising number of the principals are things that we did for Doodles where applicable for the Google assistant in a huge project and there's a lot of a lot of pieces of the But we think it's an important part of the user experience to have a set of the characters.

each of you talk a little bit about how technology interacts with humans and vice versa and how those two things have to kind of coexist so good design and thoughtful design is is is a means to make technology in this case more approachable and and useful and usable in Friendly and and to make people comfortable with that and you all approach your your work and problem-solving in this way from a very human perspective write a very like you inject and pussy it was real and talked about like humanity and

empathy injecting is empathy into into your process. So let's talk about that. Do you think can can the work you do with machine learning allowed a machine to express art in a in a human way? Yeah, let's start there with some with some constraints on how this all works. I think. What what we realized early was that we need at least two players in this game. So to speak part of the work is is building new technology. So it sometimes work we're taking on the role that a luthier might take on and building a guitar or that someone doing a music tech

program might take on a building a new kind of electronic instrument. And and I think there's a thought process that goes with building something like that. That is very creative. But I think you're also in some very real way constrained by the active building the thing to understand it in a certain way like it's your baby you built it and so, you know, you wrote the operating manual so, you know what this thing is supposed to do and in most cases what we see is that for something to become a truly expressive artistic device. It has two in some very real way be broken

by someone else and I think it's almost impossible for us as the Builders of that. Vice to also be the ones that break it and so are dream and magenta is to connect with artists and musicians and people people that don't know how to code people that don't necessarily even care much about computation and draw them into this conversation. And so what we found is that, you know, we started by releasing machine learning models in open source on GitHub, that's part of tensorflow with instructions. Like please run this 15 line long python command.

It's going to be great just run this command and then just wait because you're going to get a hundred MIDI files in a temp directory somewhere on your machine right here. He's like that's not how I make music. So what we've seen is that part of our work even even on the technologist side even as luthier's so to speak guitar maker is part of our job is to do good design and to build interfaces that people can use and then Hopefully the interfaces are flexible enough and expressive enough. Then it's a very meaningful way people can also do some fun breakage and

getting there requires a lot of moving Parts a large component of which is is very good design that notion of breaking things. You told the story once about are you made an analogy once about the electric guitar? I think and how that's is a little bit similar right like descendants that people create with with electric guitars is not that's right. I tell the same stories and let you know that Grandpa, you know, everybody was invented to be allowed acoustic guitar to overcome noise on stage and the worst, you know, they were trying

really hard to not have these amplifiers to store it to imagine a world where amplifiers don't distort and electric guitar sound like acoustic guitars. You have to have it moved very far and the breakage there was actually having fun with the Distortion and NFC going for sounds Aren't like an acoustic guitar. Connect to Isabel. I think it's so interesting about your work and that that people have to cohabitate with you physical things is that it's it's just as important or maybe more important

how people feel about these things rather than just what their utility is. How do you what kind of consideration is you make for the super starting to sound like hippies now that like help with people's feelings and their and their empathy is in there, you know the way they co-exist in the space of these things. I think a good tool that I used a lot is that I put stuff in front of people and ask them. What do you think? It looks like it's a fun game. You don't always get back. Will you want to hear but it's a really good way of testing is what the object you've

created. Does it have positive connotations or negative connotations. The first time I showed a prototype of mini I showed it to French person and I thought that was amazing because first of all, I love my mom and connotative something sweet and delicious is just excellent. And again, we surround ourselves with food, you know, that was that was just I knew we were onto something there. And in food is something universally appealing family. So so that's one exercise out of out of many. I think I think He is just to really make the thing

real really quickly to translate the big idea into something tangible and then ourselves living at living with it for a while. And then also think about you know, not only the food analogies but also making sure that the objects with design are understandable. You understand what it is. So again with many you wanted to look a little bit like a speaker in a little bit like a microphone but not too much of either but it would be very honest of that function and then connotate that this goes in the home and therefore, you know, the the fabrics and the

things that we used to surround ourselves with has to have that human touch to it. And when you find new Solutions a lot of the times they enhance the function or or help with the function like fabric is his excellent material that is in a most of the time. Audio transparence you can have life through it. You can kind of crate this calmness in the object Itself by getting all the functionality through it and I'm really passionate about trying to design pieces of technology, which hopefully people think about their just

stuff and not as technology, but that can live out in the open. There's just way too many of the furniture that are purely designed to hide technology. So my goal in life if we can get rid of those things. And Ryan you that sort of human touch is pretty evident in most everything that you do so we can talk about the the Google assistant again, it was designed to operate and to be used through the power of conversation, which is a fundamental human interface. I

guess I'm through the course of of your work on on creating a personality talk a little bit about how you found in a how you sort of steered through the land mines of What kinds of aside from the transactional things what kinds of things are people going to want to talk about with your assistant? So I think we're still steering but for us a guiding principle if our success is You did it doesn't feel like thing like this is feel like something a character. Do you want to spend time with like I missed you

earlier as far as like finding things that people we wanted to steer clear that mean is really interesting to look at the different queries, if people ask Google search and then we will ask Google assistant and a Google as you might imagine. There's a lot of people that had the background like information retrieval and like data ranking and things like that search making things like that and it kind of turned things on their head when they were asking questions, like what's your favorite flavor of ice cream or like, you know, did you fart and those are like pretty

more common than you think when people get first piece of technology that's been lovingly crafted. They also have a very different relationship to it. Very sizable number of the queries that we get on the Google Assistant like the first date queries like you have any brothers or sisters? It's really the only person looking at what is her favorite flavor of ice cream. I'm sure everybody wants to know this is very Illuminating question. Thank you for asking. So basically we have a principal. Do you know when this is Big Sur by just question to like we basically set of principles example, we

have one principle that we want to talk like a human voice at the interface, but we don't want to pretend to be one. So if you were to ask a question, like what's your favorite flavor of ice cream we would do what we would call an Artful Dodge and we look to our training and improv theater where we going to deny the user of like, you know, I do not eat ice cream. I do not have a mouth like a really bummer answer if you're exploring a new technology, you know that to shut down the conversation. I'm so we would put the same time. We don't want to lie and say like well, you know salted

caramel obviously which is like, you know a position that is disingenuous because it does not eat ice cream. I'm so we would say something like You can't go wrong with Neapolitan. There's something in it for everyone and we would take that question understand that the subtext is like I'm getting to know what you are and what you're hit abilities are and we would yes and and we've got to continue to play the game and use it actually is an opportunity to make a value statement that we are inclusive everyone everybody to the you know, that is good for everyone is good. How much

how much dialogue goes on when you're within your team and you're trying to you know, when you're talking about? Okay, what if someone asks the Google Assistant do you fart will that that was text you as soon as I that question I knew was going to be answered and it wasn't just going to default to answer. I knew that we already won like the humanist amongst us already won because there was a school of thought that you would say, I don't fart. I don't have a body and that was like and his story and that just seems that's true. But kind of not in line with you know, do you know keeping those

keeping the game going? So we would have a lot of back-and-forth and we would then like take that answer and we table at least you can say I don't have a butt cuz at least you're that you'll be a little more specific. Yeah, but but in our case we ended up with something a little more playful it a little more addressing the subtext which is of the school of like whoever smelt it dealt it which we said you can blame me if you want. I don't want How many times has that kind of you know if a user is asking about that and let's just take it a step further and put them on the spot to go ask

the Captain America use case for let's talk a little bit about how I like Google so Isabel, is that how many you mention needed to be both a speaker to microphone as well as an assistant behave like an assistant. How do you give your starting from those kind of varied engineering kind of product requirements. How do you go from there into the idea of of personality of the brand in Ryan's case his his work talks, right. It's the personality comes through that way and your work it comes through as the materials in the things. How do you consider? The personality of the

Google brand in the work that you do. Yeah. I mean, I think it's a huge responsibility and Hardware that people actually put down money for and you know, the brand is just really incredible but we're trying to figure out like what score to Google and how do we translate that into physical form and some of the sometimes it's not about like a direct translation because most people, you know don't want to pay money for something quirky. So taking that kind of principal in that idea and then thinking about what it means for Hardware I said, so in

this case for example to me Google stands for sense of optimism and kind of his optimistic outlook on the future. So if I can do things that remind people of that or that makes people smile, I think that that naturally then feels like a Google product. So just one simple example of that as you turn Upside down there's a pop of color on the back. And only you is it is a person who bought the product know that but you know, it's kind of has that Google on the inside because over the years. So for the seven or eight years or whatever however, many years you've been

to those sort of moments of delight and those who have little user experiences that that are like turning over the mini into finding like a little surprised. So everything from your responsible for the Pac-Man which is the little the little character that you drop into Google Maps when you go into street view and we talked about the personality of the Google Assistant a little bit and then of course the Doodles taking over the homepage so over the 12 years that you've been kind of working in that territory

and is the Google brand grown and evolved. Have you found how is that gross? The brand impacted the work that you do. I think the kind of Korra what I try to do. I almost have discovered it by accident like this to do Pac-Man's story for another day. But but I was just glad that I work in a place. I had free strawberries when I got here. That was very exciting to me and that they paid me to draw and be creative was just like beyond my wildest dreams on just like happy to be here. It's still am happy to be here. But what what kind of works for me because it was

always sort of my Mo was how can I use my position of privilege to bring other people up and to give them a sense of belonging and that's that stay consistent. So I know whether it's like trying to make sure we have inclusive Doodles or you know, creating an opportunity for a little like mannequin that can be dressed up for holidays or whatever for street view. It's you know, there's been a through line where it may be in the beginning Google was more of an underdog and how Google's like very important part of people's lives. I don't think you could really Play the

small organization by any stretch, but there are still human touch points that matters and make people feel like they belong which is what to do for everyone. I want to make sure we leave time for questions if anybody has them. So if you have questions, we start coming to the microphone. So we kind of go a little bit into the future future. I'm so if we're sitting here a year from now or a few years from now. Doug what do you expect to that? What do you expect that in machine learning might do for art in the future whether it's in your aspirations for the next 12 or 18

months or maybe 5 years from now. if I think that it really interesting way to think about this is considered generative models as a family of 5 Mishima new models models that generate new instances of the of the date upon which they're trained where I see us going is actually very heavily integrating the design process of products with generative model so that the we're seeing machine learning generating part of what we're trying to do and I think that's going to touch our lives in the Arts and Music and communication in a number of ways and

to those of you who or anybody in the room with developer easy question, right? Because we're at a developer conference, right? So we're going to have a responsibility as at machine learning experts to understand a little bit about design a responsibility as back-end engineer to understand a little bit about machine learning and assigned like I think we're going to see much more of a need for end-to-end integration for me the future started happening already in a sense. I have teenage kids and I watched just how that use Snapchat to communicate and how they built their own kind of grammar

around it and it's a very very Apple product now imagine 10 years of advances in assistive writing. So, you know, you're using Google Docs and you're riding and you have some machine that I am helping you communicate right? We're going to get very good at this very fast. And I expect that, you know when my kids were younger the teachers were all worried if they use Wikipedia too much to write their papers and now it's going to be like wait how much actually write like like what part of it did you write and what part of it did your assistant? Right? And I think you know, we can be dystopian

about that. There are some very difficult issues here, but it's all so wonderful. I think you know, as long as we use this to communicate more effectively and in different ways and we making it as something creative. I think it's very exciting to think about how machine learning can become really more deeply integrated in in the process of communicating and again at that's what that's what I see the Arts as being about a music thing about it's about communicating. It's about sharing our thoughts or feelings or beliefs with each other and I'm seeing In my career that happening more

deeply with machine learning as well says that's my teacher Vision Vision. What about what do you want to see in the hardware in the next year or two? Number one? I hope people find out about it smell like fish in in Milan a couple of was launched last year a lot of people to come up to me and these concepts are great. But I hope we just continue to design for everyone and everyday life and Ryan. What would you like to hold? You say it? Would you like people to take away

today? I think just remember that technology is for people first and foremost. So just always keep you know that question the back of your mind. I'd like how and what I how is what I'm doing helping people. Okay. Let's do math questions.

Cackle comments for the website

Buy this talk

Access to the talk “Design, machine learning, and creativity”
Available
In cart
Free
Free
Free
Free
Free
Free

Access to all the recordings of the event

Get access to all videos “2018 Google I/O”
Available
In cart
Free
Free
Free
Free
Free
Free
Ticket

Interested in topic “Software development”?

You might be interested in videos from this event

September 28, 2018
Moscow
16
159
app store, apps, development, google play, mobile, soft

Similar talks

Daphne Luong
Director of Engineering at Google
+ 3 speakers
John Platt
Software Engineer at Google
+ 3 speakers
Rajen Sheth
Senior Director of Product Management at Google
+ 3 speakers
Fernanda Viégas
Research Scientist at Google
+ 3 speakers
Available
In cart
Free
Free
Free
Free
Free
Free
Krishna Kumar
Product Manager at Google
+ 1 speaker
Mariya Nagorna
Senior Technical Program Manager at Google
+ 1 speaker
Available
In cart
Free
Free
Free
Free
Free
Free
Chris Kelley
Design lead in AR/VR prototyping at Google
+ 2 speakers
Elly Nattinger
Software Engineer at Google
+ 2 speakers
Luca Prasso
Visual designer at Google
+ 2 speakers
Available
In cart
Free
Free
Free
Free
Free
Free

Buy this video

Video

Access to the talk “Design, machine learning, and creativity”
Available
In cart
Free
Free
Free
Free
Free
Free

Conference Cast

With ConferenceCast.tv, you get access to our library of the world's best conference talks.

Conference Cast
558 conferences
22059 speakers
8190 hours of content