Today, one out of every seven people identifies as having a disability. When
When it comes to accessibility features in technology, things we may take for granted can have an enormous impact for others. Consider captioning.
Another key improvement in prioritizing more equitable technology, specifically for the low vision community, is the
From standard captioning to the
Of his work on accessibility features with Google, Vince says, “My proudest moments are the ones where if [the person] didn’t have it [a certain feature], it would just mean a dealbreaker.” Vince reflects in this episode that “those cases definitely warm my heart.”
Tune in to the
Transcript
[00:00:01] Rachid Finge Hey there, it's Rachid Finge. And welcome to the Made by Google podcast. Well, time flies. We've been on a journey meeting Googlers that work on our devices and services, and this is the final episode of this year. But I do have some news on that, so stay tuned until the end of this episode. So today we're talking about accessibility. It's a topic that, frankly, not enough people care about, and perhaps just as many people don't really know what it is. In a nutshell, it's the stuff we add to our products to make sure more people can use it, even if they have challenges with, say, speech, vision or hearing. There's a lot of cool stuff going on in that area, and I think that even if you don't identify as having a disability, some of the accessibility features can still be of great benefit to you. We're going to learn all about it in this year's final episode of the Made by Google Podcast. My guest today is the group product manager for Accessibility on Android. Please welcome Vince Wu. Vince Wu. Welcome to the Made by Google Podcast. Great to have you.
[00:01:12] Vince Wu Oh, thank you for inviting me. Great to be here.
[00:01:15] Rachid Finge Absolutely. So you're a group product manager for accessibility on Android. Could you describe to us what that means?
[00:01:21] Vince Wu Yes. So I guess like a lot of folks who get into this area, there's a little bit of a family member or friend. And that's how usually folks get into this area looking at how we can make our services, our devices better, serve folks with disabilities. And I work specifically on the Android platform.
[00:01:42] Rachid Finge All right. And we'll get to that in a moment. But listeners of this podcast will know that sometimes I'll go into that sort of corporate directory where we have all of the Googlers. You can see where they work, you can see their profile picture. You can also see their mission statement. So, Vince, I looked up yours and it says, Android is useful for everyone, regardless of ability, language or region. Now, I think with some people I have to ask, what do you mean with your mission? But yours is pretty clear, actually.
[00:02:09] Vince Wu Yeah, I think so. And that's what got me very interested in this area in the first place. We really want to build our products for everyone. That's why I joined Google. And when I got in touch with this project in the first place, it was a really easy decision for me to work on it.
[00:02:25] Rachid Finge Now, we've been in touch many years ago when we launched the Google assistant in Europe. But what I found interesting on your page, it also tells you the tenure. So how long someone has worked at Google and I just couldn't help but notice that you've been with Google for over 16 years. As a matter of fact, you joined Google before Android was a public operating system by Google in 2008 because you joined in 2006. What a ride it must have been in 16 years’ time.
[00:02:52] Vince Wu You know, I never thought I would be at Google for this long, but I guess one thing led to another. What has really helped is working at Google. Google has been very supportive of different regions, different roles, and every three, four years I've really done something quite different. Currently I'm based in Switzerland. Before I was in the U.S. and I spent some time in our APAC offices. And so it really feels like many, many jobs that I've done, even though I've been at Google for one stretch. One thing that connects a lot of my work is, I've been on Android and Chrome for the vast majority of my time at Google. And so that sort of set me up. Well, that gave me the confidence to work on something like accessibility, which is really a horizontal cutting role across a platform.
[Meet the Googler]
[00:03:40] Rachid Finge My guest today is a veteran at Google. I met him years and years ago when he was still working on the Google assistant. But now Vince Wu is the group product manager for accessibility on Android. In many ways, accessibility is the great equalizer. It makes sure many more people can use Android, even if they have challenges with seeing, hearing or speaking. Accessibility is one of those areas at Google that benefit so much from AI. And as Vince will tell you, we're doing things today that were unimaginable just a few years ago. I loved Finn's passion when talking about his work because it makes such a profound difference in the lives of many. We're even going to do a live demo of an amazing feature on Pixel seven called Guided Frame. I hope you'll enjoy our conversation. So accessibility, I think for some people that's maybe like this menu option. They never really checked out on Android. Could you explain to people what does accessibility mean or what do we mean by it?
[00:04:46] Vince Wu So we really mean making sure that the devices work well for all users. Literally, we want to build for everyone, regardless of their ability. And so I think the most obvious thing that people get connected to is, well, we all know folks who may have a disability. This could come in many forms. It could be a vision related disability. It could be a hearing related disability, it could be motor, which relates to dexterity. Whether you can use fine motor skills, it could be something else. There's a broad range of this, and one of course makes sure our devices, the growth of devices in our lives. That we all, we all have smartphones these days and maybe even a tablet media watch, maybe even your TV, your collection of devices that you have in your home are all smart and connected these days. And we want to make sure that everything that runs on Android works well.
[00:05:37] Rachid Finge So any Googler I spoke to in this podcast, or basically every Googler I ever spoke to, they're very passionate about the work they do. So where do you get the passion to work specifically on a topic like accessibility?
[00:05:50] Vince Wu Yeah, for me I think it's very similar to a lot of folks on accessibility, which is often we have a family member or we have a close friend where we've directly seen that's where they may have struggled with using devices. And I think technology for me. There's like two sides of it right on one side. The more devices that we have, the more we rely on devices for services, for being able to access things that become a barrier. Right, if you can't use it. So that's one side of it. And then the other side of it, though, which is very exciting, is there's a lot of things that we could do with the power of technology that was not possible in the past. I think it's very fascinating to see both sides of it, and it's very interesting to be in the middle of that.
[00:06:35] Rachid Finge So maybe we can make this a little bit more concrete. Let's say someone has issues with their hearing ability, for example. What is something you worked on in Android to help them continue to use Android?
[00:06:49] Vince Wu Yes, I'll maybe jump straight into something and then maybe we can broaden out just to get a better understanding. Well, the first thing I would mention would be captioning. Right? So I think a lot of us take captioning for granted these days. A lot of times. Generally folks might turn it on. Maybe if they don't want to listen to the audio, you can do it in many different languages. It works across many things. YouTube videos. You can get it for phone calls, you can get it for meetings or everything. But I think if we zoom in to the need of someone with a hearing impairment, you of course, won't be able to hear a lot of the things in your life and being able to caption things, be able to understand the sounds around use of captioning, not just with text but more broadly around like, hey, what are the sounds around you? Maybe there's a baby crying, maybe there's an ambulance close by. So I think all of that comes into it. And I think if we look at things a little bit more broadly, what we'd like to do is look at the different types of users in a range of disabilities that they have. So for the most extreme, let's say someone can't hear anything at all. Something like captioning not having it would be a deal breaker. Absolutely right? And then if you look at someone maybe with just a little bit of an impairment or maybe it's just situational where it is not easy for them to listen to something maybe not convenient, then captioning becomes something that they turn on and off. Maybe they turn it on for an hour because they need it and then turn off later on.
[00:08:22] Rachid Finge That makes sense. And then I think as you alluded to, there are many people you can help with the accessibility services that Android has. So for some people might be hearing, others might be vision. You mentioned dexterity as well. How do you decide what to work on? To me, it seems like everything is important in a way, but yet there has to be some sort of priority. How do you tackle that?
[00:08:46] Vince Wu That's a great question. I think as people who gravitate towards this field, we definitely want to work on everything. Right? Because part of working on this is you start with the user and we meet with people and people have different needs. And I think coming into this, you just want to help everyone. I think there's some things where there's existing technology and you sort of take advantage of that. There's certain things where there's new tech coming in and you sort of see that, Oh, if you look at the trajectory of that, it may make a big difference if we take advantage of that and leverage it and jump on that. So I think looking at where technology is going, where technology is heading, and then seeing what kind of things could be the most helpful. Maybe to give a very concrete example here. So recently we looked at the trajectory of general image understanding, and so we saw that. So this is for folks with vision impairments and they can't see or they may need assistance with seeing. So you want to connect them to the ability to understand images or understand icons, you know, text, that sort of thing.
[00:09:51] Rachid Finge And then the phone will tell me what it sees through its camera, for example.
[00:09:55] Vince Wu Yes, the technology that was very mature was around icon detection. And so we jumped on that. And the example there is that if the developer, when they're building an app and they forget to label certain icons, for example, you have a home button, a back button front, but if we forget to label it, then there's no way for the end user to know what they're navigating towards because they're using a screen reader type of interface. Right. And so with icon detection, we can then figure out, oh, we know this looks like a home button. We can tell the user that's what it is.
[00:10:27] Rachid Finge You mentioned that obviously you're talking to people with disabilities to understand what they actually need. I guess that falls at the research department. What's that like to sit in such an interview and learn these kinds of things?
[00:10:41] Vince Wu Yeah, I think the difference between this project and other projects is I always learn something new on this project, whereas on other projects, maybe after a couple of months, I feel like, okay, you know, I pretty much, I understand a user just like different groups. With accessibility, there's just so many different groups. And there's also intersections of disabilities, like, for example, someone who's aging, getting older. They have a bunch of things going on all at the same time, maybe progressing at different rates. There's a difference between developed markets like the U.S. versus other markets that maybe are still developing, like, for example, India. I think there are many, many different angles to this. Some folks may have access to expensive devices like hearing aids, and some folks may or may not depending on the situation and the user segments and their needs. There's just a ton of diversity here.
[00:11:39] Rachid Finge Absolutely. Now, we've talked about captioning, which will then in the end rely on AI and machine learning, a topic that frequently comes up in the podcast. What kind of role does I play in accessibility and did the field change because of the advances in AI?
[00:11:57] Vince Wu Oh, absolutely. I think I would struggle to find an area where AI is not touching or is not going to touch, because just by definition, we're talking about people's senses. Right? So your eyes, your ears, your motor skills, may be degrading or you might have a need around that. And that is exactly the fundamental research that is going on in A.I. to mention a few other examples other than captioning. So we have sound sensing. That's another area in the hearing space. In vision, we have image captioning. We have an app called Lookout, which I really love. This is an app that not a lot of people are aware of. But what's cool about it, is that for people who may have low vision, they can open the app and there's different modes, right? There's different types of things that it can detect. You can pull out some bills from your wallet, and then one of the modes is identifying different kinds of bills. Another one might be food. So if you go to a grocery store and you look at the labels that you see on boxes, it'll tell you what kind of food you're looking at. And yeah, that's just fascinating. The types of things that you can detect.
[00:13:07] Rachid Finge Look out. I guess that's just in the Playstore for everyone to download.
[00:13:10] Vince Wu Yeah, yeah. You can. Just Lookout, one word here.
[Made by Numbers]
[00:13:13] Rachid Finge Amazing. Now, Vince, we have a recurring segment in the Made by Google podcast called Made by Numbers where we ask our guests to bring a number that's either important to themselves or to the work they do. And we've had all sorts of numbers from very tiny ones up until millions and billions. So just wondering what the number is and maybe I can guess what it relates to.
[00:13:33] Vince Wu Okay. One in seven and one in five.
[00:13:38] Rachid Finge Oh, that's interesting. So you're the first guest to have a sub one number. That's nice. One in seven and one in five. I'm thinking that maybe one of those two pertains to people who identify as having some sort of disability.
[00:13:52] Vince Wu Good guess. So that is one in seven. One in seven people in the world have a disability.
[00:13:56] Rachid Finge Right. Okay. But then the one in five, I would not be able to figure that out.
[00:14:01] Vince Wu Yeah, the one in five. Maybe it’s a little bit of a trick question because it's not entirely related. One in five is more around the idea of the work that we do is actually applicable more generally. The one in five is the number of people who are older than 60 by the year 2050. Because I guess you know that the world is aging, right?
[00:14:21] Rachid Finge Sure. Okay. And so looking through those numbers, that means that maybe in 2025 or 30 years time, your accessibility features might be more used or will have more popular demand, I guess.
[00:14:34] Vince Wu Absolutely. A couple of weeks ago, I was actually at my parents’ house. And obviously I like looking at their phones and like seeing what they're seeing, seeing how I can help them. And I show them a few features that historically has been considered accessibility, like things like changing your font size, changing the color contrast, maybe some magnification features, and they don't need it all the time. They might need it maybe a couple of hours out of the day when their eyes are a little bit more tired than usual. But what they saw there was, Oh, this is cool. Like I never thought to use it. And then they loved it. And now they know how to turn it on and off. Right. I think this is one of those examples where things are more relevant and than maybe they seem and they will become more relevant over time.
[00:15:16] Rachid Finge Now there's this new cool feature in Pixel 7 and Pixel 7 Pro that launched last month, which is Guided Frame. And rather than me explaining what it is, maybe you can tell us what Guided Frame intends to do.
[00:15:28] Vince Wu Oh yeah, yeah. Very proud of that feature that we launched. So it really came from us really trying to look at things from another point of view. So historically, a lot of teams, hey, how can we just make sure that things are accessible? That's sort of like the starting point. And then we really try to pivot away from that, like, hey, let's look at what people are trying to do with their phones, right? And then figure out like I think the best that they can be. And so when people use smartphones, what do they do? One of the top things is, they would take pictures. And so once we started looking at it from that angle, it becomes very obvious. So we looked at people with vision problems and it was actually quite interesting at the beginning, people. Like, Oh, do we have to look at this? There was a general comment just for folks who were not in this area is like, oh, do people even want to take pictures? And then we got a lot of clear user feedback and we shared it. It was like, Hey, obviously everyone wants to take pictures and people also want to take selfies. You just want to they'd have a phone. They want to do that. And then we were looking at the experience of doing that, and it was really hard for people to get a good selfie because you kind of need them to frame yourself right in the right place, of course. By the time you like, press the button to take the photo, the motor moments of your fingers actually changes the angle. So people thought they were taking a good selfie, but they're not.
[00:16:44] Rachid Finge And then Guided Frame can help with that. And this is going to be a first in the Made by Google podcast events because we're going to actually do a live demo of this. All right. So I enable Talk Back on my phone. Talk Back is what many people use in Android when they have a vision impairment. And then the phone kind of tells you what's on screen. And once you have Talkback enabled, that will enable Guided Frame when you're in the camera app, basically. Right. Yeah. So I'm going to unlock my phone.
[00:17:13] Phone Switch to camera mode. I'm Auto Max. So you're okay. Says path.
[00:17:18] Rachid Finge Let's go to the selfie camera.
[00:17:20] Phone Ready for one cropped face. Move your phone left. So no faces in frame.
[00:17:25] Rachid Finge I'm mis-framed, apparently. Let's see if I can change. I'm closing my eyes, by the way. Yeah, that's like.
[00:17:32] Phone One cropped face. Move your phone left. Okay. No faces in frame.
[00:17:37] Rachid Finge Or the other left.
[00:17:38] Phone Sorry. One cropped face. Move your phone. Left and up.
[00:17:41] Rachid Finge Up. Okay.
[00:17:43] Phone Ready for selfie. Three, two, one. Photo taken. One face. Good for selfie.
[00:17:52] Rachid Finge That's actually amazing, Vince. That actually tells me or the person who needs this how to frame themselves.
[00:17:59] Vince Wu And you felt it you know you felt it directly with your eyes closed. How hard that could be, right?
[00:18:04] Rachid Finge Yeah, absolutely. It is really hard because I actually felt like I was doing a good job and it was still coaching me off that you should turn your phone a little bit more to the left. That's amazing. And I think for many of us listening, it's so first nature by now to take a selfie. And then now since Pixel 7, you're enabling a lot of people to maybe do that properly for the first time.
[00:18:24] Vince Wu Yeah, something that doesn't come through in the audio is that we tuned a lot of a combination of audio and haptics, so you must have felt it in your hands when you were getting closer. It would give you that haptic feedback and it's not obvious. But we went through many, many, many iterations of that and tested it at so many times to just make sure that it would work.
[00:18:45] Rachid Finge That's amazing. And maybe we can talk about some other features that are there for people to use in accessibility. So we've talked about captioning. We also have live transcribe, which is sort of similar, but it's not quite the same. Could you explain the differences?
[00:19:02] Vince Wu Yeah. So Live Transcribe is for more of a conversational mode where let's say you're having dinner, right? And then you have a family around you and you want to make sure that everyone participates in the conversation. Then you would pull up Live Transcribe, and then you would be able to see everyone's voices transcribed into text for you.
[00:19:24] Rachid Finge Right? So at the dinner party, I put my phone on a table and I'm going to see what everyone's saying.
[00:19:30] Vince Wu Exactly. Exactly.
[00:19:32] Rachid Finge All right. And you mentioned the Lookout app, but it's easy to understand why this is useful for people. If you can just tell me what I'm seeing. And again, this is something like if you were thinking maybe five years ago when AI wasn't as prominent, this would have been close to impossible to make, probably.
[00:19:49] Vince Wu Absolutely, man. You were asking me earlier around like, you know, hey, I joined Google or like, you know, 15, 16 years ago, I would have never imagined that we would be able to do something like this, that this is all advances that just came about in the past, I don't know, like seven or eight years. And things have progressed really, really quickly. And just looking at the future, I'm really, really energized by what is possible.
[00:20:13] Rachid Finge Well, something that, you mentioned before, but something that to me sounds like from the future. But it's actually here is sound notifications. What is that?
[00:20:23] Vince Wu Yes, sound notifications. Let's say I'm in my home and I'm hard of hearing. And but I do, of course, want to know if there's interesting things happening. An example might be like my carbon monoxide thing is beeping. Maybe I have things in the kitchen that are beeping that I would want to know. So things like that.
[00:20:43] Rachid Finge Right? So would I then sort of make a pre-recording of it and then it will recognize it in the future.
[00:20:50] Vince Wu Yeah. So right out of the box, when you have the service, it will detect things in general for you. So this is a pre-trained model that we've done based on a lot of sounds that we've collected, but you can also customize it too. So we recently added an appliance customization feature, so we call it custom sounds, so you can record the beeping sound of something in your home that might be a little bit different, right? And then you can label it your own. I actually have a kettle that emits a slightly different sound, and so I did it for my phone.
[00:21:22] Rachid Finge Right? And then when the kettle makes that sound, what happens then on the phone? Yeah.
[00:21:26] Vince Wu Then you would automatically get a notification and you also get a history of you as well, which is really neat. So you can go back in time and say, Hey, what's been happening? Did my dog bark a couple hours ago? Right? You can see what has happened around your home recently.
[00:21:41] Rachid Finge Right. So this is going to be a tricky question to ask because you just said it 16 years ago. I would have never thought we would be able to make something like we have now. But I still need to ask you about the future. What's in the future of accessibility? What's maybe something you'd love to tackle or to build that maybe isn't just possible today?
[00:22:00] Vince Wu Something that comes to mind is this, I've been thinking about it a lot is sign language detection. So what's possible today is we can detect pretty quickly like if people are like typing with sign language, right? Their ABCD, right? But general purpose understanding of sign language and making it superfluid where it's kind of like the transcribe example. We're conversing, maybe we're in a video call together, let's say I only know sign language and maybe it's what's I'm most comfortable with. And then you speak back to me in some language and could be any language. So imagine sign language is being one of the many languages that are supported and we can go back and forth. You talk to me in German, I speak to you back in American Sign Language. Wouldn’t that be neat?
[00:22:44] Rachid Finge And then the American sign language will be converted for me into. Well. German speech or text. I guess I would speak in German, which. Yeah, why not? I shouldn't do because it would be terrible. But if I do that and it would convert it back to you in I guess sign language graphics or something like that.
[00:22:59] Vince Wu It could I credit it. Yeah, you can definitely render the images and do it that way.
[00:23:03] Rachid Finge That's amazing. And this is again something that maybe five years ago was unimaginable that you'd even think about trying and build something like this.
[00:23:11] Vince Wu Yeah. Now I feel like I could have a conversation and maybe people wouldn't laugh at me.
[00:23:15] Rachid Finge Right. And I'm not laughing at you. I just hope you do it someday. That’d be amazing. So. Vince. One of the final questions I have is I'm thinking that at some point you run into people that have a certain impairment or a certain disability and they find out that something you worked on actually helps them use technology, whereas previously they couldn't. What's sort of the proudest moment where you saw someone using an accessibility feature you worked on?
[00:23:46] Vince Wu I think my proudest moments are probably the ones where if they didn't have it, it would just mean deal breaker. Right? Those cases definitely warm my heart. But I do have to say this. I'm a product manager. I gravitated towards, hey, what's what's not working for you, right? And tell me the bugs and I'll go back and fix them. That's maybe a little bit of a job coming into it.
[00:24:08] Rachid Finge Got it. And is there any specific memory like someone you ran into or maybe in a research group or maybe it's a friend or someone in your family where you had an experience where you undid the dealbreaker, basically.
[00:24:19] Vince Wu Yeah, surprisingly, one of my friends didn't know about captioning and I let them know about it for it's a call specific feature so you can actually make phone calls and then the captioning would come up and I guess it just wasn't well known and I showed it and it was basically a one of those dealbreaker moments.
[00:24:40] Rachid Finge That someone who previously couldn't do a phone call and now was able to do so.
[00:24:43] Vince Wu Yeah, or they might have had to ask someone to help and it's possible. So it's not like completely dealbreaker. But this is about independence, right? Sure. People want to be able to do things themselves and that is very, very important.
[Top Tips for the Road]
[00:24:57] Rachid Finge And that's a good thing for people to know that live captioning actually captions, everything, including phone calls. So yes, yeah. That's definitely a great feature there. Vince, we've come to this segment of Top Tips for the Road where we ask our guests Top Tips for people to try out. Or maybe it's Top Tips about being a great product manager or any of these things. It's basically your moment. What are your top tips for our Made by Google podcast listeners?
[00:25:26] Vince Wu I think what comes to mind is maybe the broadly useful accessibility features that maybe aren't, maybe they're kind of hidden, right? So the most useful ones broadly applicable I think are things around like font sizes, color, color contrast, magnification. I use this one a lot called a selective speak where you can select text. Mm hmm. And then you can just tap it to for to speak it back to you. And I find it very useful because sometimes I'm just not in a situation where I can just be like paying full attention to my phone and I use it.
[00:26:03] Rachid Finge That's amazing. Vince, thank you so much for your time. We learned a lot about accessibility today and the very important work you and your team is doing. So thank you so much for joining us.
[00:26:14] Vince Wu Thank you for having me.
[00:26:17] Rachid Finge I feel very inspired after talking to Vince. Everyone should benefit from technology and I'm proud that we're working really hard at Google to make sure no one is left behind. So two highlights from Today Guided Frame available on Pixel 7 and Lookout is an app you can download from the Google Play Store, so go check that out. We hope you enjoyed this first season of the Made by Google Podcast. We had a ton of fun making it and learned a lot along the way. We started with Isaac, who works on Pixel Camera and explained how he wants to create more confident photographers, by which he meant more confident photographers and more confident photographers, if you see what I mean. We met Jesse and Ray telling us how they build Android and Pixel pretty much like a castle to keep your devices and personal data secure. Isabel told us all about design at Google, talking about, for example, Nest Wifi Pro, why it matters that it looks beautiful because it also helps you get better wifi and how manicuring your nails in a certain color is a great way to test that color. Monica explained what makes our tensor chip different from other chips. Hint: It’s AI, of course, and why synthetic benchmarks only tell you part of the story and usually a part that isn't very interesting. Nino from the speech team taught us a lot about how computers learn to understand human speech and how perhaps in the future we could tune the system to the way you specifically speak. DeCarlos works on Fitbit and Pixel Watch and told us what all those sensors do in the device and. And where pixel watch is heading next in terms of helping you improve your health. Karen and Anish explained why matter matters so much or what's the matter with matter? All those puns. But matter is the new standard that will help us bring your smart and thoughtful home to the next level. And then today we had Vince showing us that AI is beneficial to everyone, including and perhaps especially if you have a disability. It's been a true privilege to meet all these amazing Googlers and bring our conversations to you to shine a light on the things we're doing at Google to help you make your day just a bit easier and more delightful. So is this the end? Well, I have some great news to share. We are exploring great new projects, interviewees and numbers for a second season of the Made by Google podcast. So in other words, do not unsubscribe or actually do subscribe if you haven't yet because we will be back. Thanks again so much for listening. We've had a lot of fun making the podcast and we hope you enjoyed listening to it. Let's meet again next year, shall we? So please take care. Enjoy the end of 2022 and talk to you next year.