Mark Webster: [00:32:36] And what I think that they absolutely nailed in this ad campaign, besides adding celebrities and sort of tapping into pop culture which I think is smart, but the one thing they nailed is that all of the commercials basically are showing the inner dialogue of everybody throughout the commercial and no one is actually speaking. And I think that the idea of the frictionless nature of a voice interface and being able to use it as sort of just like that next extension passed thought, I think they kind of nailed it here. Everybody always asked me like what is the killer Apple voice? You know like Uber existed before mobile, you know Instagram couldn't have existed before we all had a camera on our smartphone. Like what's the thing that voice will have the killer app for? And I think that the killer app for voice is convenience. I think it will just tie us into everything else we use. And I feel like this ad campaign really gets at that in a really smart clever way.

Cheryl Platz: [00:33:38] Well I'm glad you brought this up, but and it was nice to see the article that brought all these pieces together and I noticed the uptick in their spending also during the Olympics, there were so many ads for Google Assistant during the Olympics. And I think what's interesting for me looking at the videos is it's a very bold promise. I've always said when I was talking about the differences between the devices today Google Assistant strengths, their strengths are Q&A, definitely better Q&A than Alexa, much more succinct, much more accurate. And also productivity with respect to the Google ecosystem, but when you say "make Google do it" like the action implied by the word "do" is very broad, like we don't consider conversational tasks necessarily doing something so it does go a lot broader than the voice interaction. I think the promise in this ad campaign and I don't know how that will play out over time, and I also think I kind of agreed with one of the points I believed that was in the article that it's interesting that they're using celebrities. It's definitely differentiating themselves, but it's also creating this sort of other, like maybe Google's for affluent people or for people who have lives that I don't, or busier than I am. It's interesting to see how the personification of a device helps us connect with it on a human level.

Cheryl Platz: [00:35:03] And celebrities are hard to really connect with, we view them as an "other" that are fascinating to watch. So I'll be interested to see what effect this has on how people perceive it. I think Mark had some great points. It's a really great point that Mark made about how it's more about the dialogue between the people than it is about the dialogue with the device. That's really interesting because as a designer we kind of want our interface to be invisible and in a way in these ads it is, but this almost feels like the opposite of personification. It's like now Google Assistant is this abstract concept that does things for me in the background and I don't necessarily talk to it terribly much but when I do it gets it right. I think it's a solid play from trying to differentiate themselves from Alexa, and well and Cortana, but Alexa is really the major player there.

Mark Webster: [00:35:54] Yeah it's a great point about whether or not they're getting out ahead of themselves right. To Bradley, you were just speaking about it earlier. I think one of the things that Amazon got right with Alexa is gave us a promise of this voice assistant that can do like four things really well initially and we were all delighted by it. We all got Siri in 2011 and we're sort of told Siri could do anything and that wasn't really true and it just led to a lot of disappointing experiences. And for a guy who's devoting his life to voice interfaces, I never use Siri. And so it's interesting to think about like is Google now sort of going to get ahead of itself and set themselves up in a way to disappoint users in a way that I think Alexa has managed much more effectively.

Cheryl Platz: [00:36:42] I also don't use Siri so we're in the same boat here, I don't know what to say about that platform.

Bradley Metrock: [00:36:48] There are rumors that there is someone who uses Siri out in the wild somewhere but they're like a unicorn you know, they're like the lost little lobster. Brilliant commentary, but Cheryl you touched on exactly what I thought when I looked at this, and just from a very personal standpoint, I looked at the ad campaign and I thought, make Google do what you don't like because you're right. Now you have my attention. You've got my curiosity now because I'm asking, make Google do what? Make Google make my life a lot easier with email? I'm down with that, desperately need that. Make Google handle all of our meetings and calendar invites and all the back end stuff that takes a bunch much time? Okay, and you know what both y'all are saying, there will come a time you know I don't use Google Assistant very much at all, but after this ad campaign I have to say at some point I will dive in a little deeper. And you're exactly right, when that moment takes place if it is not ready for prime time or it gives me some garbage like you know that time I tried to use Siri to simply dictate a few notes and it had no idea what I was asking for. It would have been easier just to get out a stone tablet and then write it and then dictate off of that than try to use my phone for that and Siri. You know when that moment for prime time comes, she got to be ready and you're right, absolutely right. I think it's a very interesting point. Are the over promising or is it something they can deliver on?

Mark Webster: [00:38:33] I think this is all starting to blur a little bit of voice in general, right. So voice is going to continue to be a more and more popular form of interaction with all of our devices across every medium and then sort of the assisted nature and the A.I. behind it, it's sort of a separate question and sort of like what does that deliver well. So dictation versus sort of you know AI help with some task are two very different problems to try and solve and for better or for worse, all of these companies are basically taking an approach where they're baking that entire experience together and making it one in the same, right. So when people talk about can Alexa suggest a good song to play. So whether Alexa has a voice interface works well is different than music personalization which is the whole reason Spotify and Pandora exists, right. I think back to sort of whether or not we disappoint or delight users, there is the risk that we're running just from bundling all that stuff together from an expectation point of view.

Cheryl Platz: [00:39:41] That's some fantastic points about sort of conflating AI and voice URI. I think the advantage Google may have, and I haven't seen them really leverage it yet, is their multimodality, or their opportunity to be greatly multi-modal not just voice but touch interfaces and GUI, graphical user interfaces. They have an OS and it's on a ton of phones and that's always been a sort of a holy grail for Alexa as it turns out with the Echo Show and with the app trying to combine these two worlds. And they're operating at a disadvantage at Amazon, they don't have an OS. They tried with Fire Phone that did not work out. And to your point earlier, there's not a lot of talking in the ads but I wonder if this leaves the door open enough that when we do inevitably give Google Assistant another go, it takes things to the next level and uses voice as just one part of like a healthy sort of healthy balanced breakfast of interactions.

Mark Webster: [00:40:43] I mean if you just look at the ad where Chrissy and John where they're trying to find a show to watch, like how did Chromecast not play a role in that commercial?

Cheryl Platz: [00:40:54] Yeah, that's a great question.

Bradley Metrock: [00:40:56] That is a good question and that was the one I thought that was hilarious by the way, where he was singing the song about how he feels like they've been trying to change the channel for two hours. We've all been there you know, we're trying to experiment with technology and it is just like not working. Yeah, so all in all great commentary, I appreciate that.

Bradley Metrock: [00:41:20] Moving on to story number four. This is our Voicebot.ai story of the week. Voicebot.ai is a very good news and commentary site on all things voice and AI. This week we have a little bit of a different story. This is original research that they have produced and in Googling this and getting ready for this show, Bret Kinsella and all the folks with Voicebot.ai and all the folks that were involved with this study deserve a lot of praise because this thing went everywhere. This thing has been cited all over the place, in all sorts of different media outlets. And Cheryl I want to start with you. What did you take away from this? I don't really want to color it too much; I just want to get your thoughts just right off the top.

Cheryl Platz: [00:42:09] I had a reaction as a speaker who's been looking for statistics to help convey the penetration of these devices in the U.S. market, but then I also had a reaction as a designer. This seems on track for me so I wasn't super surprised by the number, especially because some numbers I previously heard said about like 1 in 10 U.S. households had one of these devices that if you sort of extrapolate out, a lot of those are couples or like roommate situations, like oh OK that's about 20 percent of U.S. adults. It sounds like things are progressing along the same very good growth track we've been seeing.

Cheryl Platz: [00:42:47] But you know one of the topics I like to speak about is about the risk of bias, especially in voice user interfaces, the way the data models are built, and the way that speech science relies on the recording corpus that we have from customers. But these devices are really only going out to fairly affluent families at this point. And so I see the 20 percent number and I kind of wonder how long it would take to bridge the gap. Like who are those 20 percent and are we at risk of sort of creating a feedback loop where these devices are only really good at recognizing the accents, and dialects, and speaking patterns of folks who have the money to spend on these speakers? I'm going to be watching this number with interest you know over the next months and probably years afterwards because if the numbers flattens and if that doesn't continue to grow there are probably voices that are unheard. And I wonder what we as an industry can do to make sure that this technology, which can change lives regardless of your income level, and in some cases disproportionately for folks who don't have access to other technology, how we can make sure it's more equitable and inclusive.

Mark Webster: [00:44:01] Yeah those are great points. It also makes me think of you know Google Assistant being in a place you know rolling up to different countries, much faster in different languages. If Google ultimately over time builds a much better set of training data for that. And yes it's a great point. Looking at the study you know the two things that sort of jumped out at me were you know, one I think maybe even a year or two ago there was a question about whether it was going to be a fad and it kind of feels like we've moved past that point. And now there's sort of broad adoption of these platforms and the two things that jump out to me. One you know as big as exciting as these platforms are, and they are exciting, it's also going to shift user behavior right.

Mark Webster: [00:44:53] So back to some early points of, if every day in my house I'm now talking to a voice interface, when I get into my car it feels weird to flip plastic switches to do stuff right. When I sit down at my computer at work to have to know all these keyboard shortcuts to do stuff, I think that this will just become the big shift in human interaction overall, even outside of smart speakers and even outside of these platforms. You know this is up there with you know the mouse, and clicking, and keyboards, and tapping on your phone. And then the other big thing was just the opportunity everybody has to you know, companies and brands to really shift user behavior. When you look at some of the numbers around like how many people pay a bill with these kinds of platforms? Like I don't know if that was any behavior that I would have thought jumped out in an early platform and I don't know if it was anything people were just super eager to do, but you know when you look at a company like Capital One getting so early to these platforms I think there's an opportunity to steer user behavior a little bit and introduce experiences that are super convenient and delightful and change the trajectory of how people interact with these things and what they do with them.

Cheryl Platz: [00:46:10] It's interesting, I think you mentioned people. It's changing the fundamental way people interact with devices not just at home, but maybe at work and work is a really - Cortana's on a lot of devices right now but you don't see the adoption in the workplace. There's still this like social barrier and I wonder where the tipping point is where people will no longer feel sort of embarrassed to talk to devices in a public space like a workplace where judgment might be a risk. Is it 20 percent of adults? Is it 50 percent of adults? What kind of usage patterns will lead us to that secondary voice tipping point where it's not just a fun thing to do at home or in my car in private, but something that I can genuinely accomplish around other people?

Mark Webster: [00:46:53] Yeah I wonder if there was that same social stigma like when phones first came out on people's desks, right? You were all at work and then somebody got a phone and then, Oh I thought you were talking to me and like you're talking about somebody that was sitting on your desk, like that definitely feels like we have all these devices around our desks. It definitely feels weird to everyone else in our workspace like when we talk them. You know it's like when do we hit that jumping up point like it's expected behavior, it's no longer weird.

Cheryl Platz: [00:47:22] And I think this may be getting a little bit off topic, but I'm fascinated by that, like how good device interaction model might change because having worked at Amazon on Alexa, it's not very scalable to have an Echo on every desk. If a couple of people forget to mute their devices, it is a little overwhelming when you ask for the weather. But as we see the Voicebot.ai story and the growing penetration of these interaction models, it's probably going to have to shift for more public spaces. It will be interesting to see how these trends influence whatever that interaction model ends up being. And the word adults jumps out at me too because back to your cameo appearance from the earlier session from Benicio, there's a generation of kids not represented by this data. And I think rightfully so because we want to respect COPA and we don't want to get a difficult line these devices are walking, but I wonder how many kids are being exposed to this and how that's going to change what they expect from the world?

Bradley Metrock: [00:48:25] Oh sure and we had a very interesting discussion on last week's episode of This Week In Voice about really touching for the first time on this show about do we need to protect children from anything or are we rushing sort of headfirst in there? Or is it better to just sort of let this tech cycle play out like they always do and not be too protective about things? I would invite anyone interested in that conversation to go check that out. We actually did something rare, we excerpted that audio from that episode and slapped it on the page for last week's episode on ThisWeekInVoice.com, so you can check it out there. It was a very interesting conversation.

Bradley Metrock: [00:49:08] Cheryl, I want to touch on something that you said right off the start. It is something I think about from time to time just with the show. You know we've shot up in adoption to where we are now. You know 20 percent adult have smart speakers. If you knew the median number of smart speaker is in every house across the world, it would probably be a shocking figure. But you're touching on, and you brought up the concept that it may not be a smooth path to 100 percent. There may be economic limitations, social limitations, and so I guess I'll ask a question. Do you believe that things on their natural path, that every home will have voice technology in it in the near future? And if you don't believe that, then what in your mind do you think - what responsibility, you know last week we would talk about parental responsibility, you touched on that. Now we'll touch on responsibility of tech juggernauts. What is the responsibility of tech juggernauts to make sure that this technology gets out there in the way that is most beneficial?

Cheryl Platz: [00:50:31] What a great question and so many opportunities there. I mean I'm a big proponent of asking why, you know why do we need 100 percent penetration? But assuming that that's a good metric to get to, what benefit would it give the human beings on the other end of that top 20 percent to have these devices in their lives? Are there problems that they have that we're solving, are there different problems they have that voice could solve? I suspect that's probably the case. We just don't know what we don't know. And I think it is the responsibility of those who have privilege, and those of us in the tech industry, we all have it in some way or another. It's been a harder one for some than others.

Cheryl Platz: [00:51:13] But I think we all have this opportunity to say like we don't necessarily understand how this technology could change people who are not like myself. I think as designers it's very exciting, because I think there's a great opportunity towards like participatory design reaching out to folks who wouldn't naturally adopt these systems and saying, are there problems in your life where you're dealing with extreme constraint, or you don't feel comfortable using your computer, or you feel cut off from the world because the technology you have is intimidating. You know I think it's interesting, there's an interesting opportunity in India I think because there are some literacy challenges for the general population that voice technology can solve. And maybe that's a place, an example, of the problems that can help us get to the last sort of 20 percent of voice adoptions.

Cheryl Platz: [00:52:02] And so I think to go back to your original question, like what is the responsibility? Always be asking ourselves questions like how does this technology affect and potentially benefit folks who are not like us? How do we extend a hand across the gap towards people who are not like ourselves? The secretive nature in which these products were launched was sort of you know capitalism requires that first mover advantage, but it meant that all of the initial testing was essentially on internal employees, and they share some very common experiences and it takes work to break ourselves out of that bubble. The good news is I think there are genuine benefits for folks dealing with disabilities, for folks dealing with the intimidation of working with a PC for something that's just sort of alien to them. But I think we just have to keep extending that hand across.

Mark Webster: [00:52:57] In general, I'm optimistic of what the voice interfaces will bring to other populations. The Wall Street Journal ran a great article about six months ago saying how the next billion users of the Internet will be driven by voice and video. You know pointing out literacy issues I think and opening it up to other populations will be amazing and transformative. You know I am very encouraged to hear a principal designer at Microsoft who's asking these questions, who worked on the Alexa team who ask these questions. I actually got intrigued by voice interfaces from looking at senior care solutions. Like we were looking at eldercare and in-home monitoring for elderly populations and got excited about voice and sort of stumbled on the problem of designing prototyping for it.

Mark Webster: [00:53:45] Our lead engineer Kamryn comes out in a world of accessibility and so we talk a lot about things like you know the breakdown of multi-modality aside from user convenience until there's a screen. People who have hearing issues can't really interact with voice platform threads. You're not getting any training data for the deaf accent and how the response would be a visual response even though the command would still come from voice right? And so when you think of accessibility issues and where multi-modality plays a role in how all that stuff will play out. And again to Cheryl's point, I think it's good that we have the right people sort of asking these questions and that company's big and small are trying to be responsible and think about how we create technology that is inclusive and helpful to everyone.

Bradley Metrock: [00:54:35] Yeah that's great and I have no doubt will be a dramatic concept that comes up on this show regularly. I mean it would be foolish to assume that we're just going to go from 20 to 50 percent to 70 percent adoption, you know all the way to 100. I mean there's going to be a little bit more engineering that probably has to go into getting it to that level. A little bit more you know intention that has to go into that. So that will be interesting to watch as the market matures how that plays out. I appreciate the commentary on that.

Bradley Metrock: [00:55:11] We will roll on to story number five. "Hey Cortana" is now "Cortana" and this article talks about how Microsoft is testing Cortana and Alexa integration that we heard about last year internally for release in the near future. And Cheryl I'm going to start with you on this, but before I do I want to say, so I'm a gamer. I believe you're gamer as well. Is anyone not a gamer these days? And I own an Xbox One X, and we have a Thursday night gaming get together that's lasted for about a decade now. A friend of mine was on and you know he did not interact with the Xbox live in any other way, or primarily with Cortana, and he had his Kinect plugged up and when I sent him a party invite he said "Cortana open that" and it opened it right on up and I was like oh my god why am I not doing that? So just to say Cortana was super functional in a way that I had no earthly idea, I want to ask you sort of your thoughts on this story, but then sort of share with us whatever you may be able to about the state of Cortana right now and what you guys are working on.

Cheryl Platz: [00:56:26] Well so I am also a gamer so it's been tough for me to watch the curve of the Kinect in the marketplace, because gosh I think that I was so in love with the Kinect in the way the voice control changed the way I dealt with the OS. And over the years we ran into sort of the branding and marketing problem right, like the Kinect cost a bunch of extra money and because the platform was perceived as a gaming platform, people weren't interested in the crazy benefits of the natural language interaction which was very cutting edge at the time. So they didn't buy the device, so we didn't hit critical mass. So now it's essentially been deprecated functionality, which is such a shame because as you point out there is some real benefits to using natural language to control the Xbox. We don't do a lot of Xbox gaming in our household. We do sometimes, but I tend to be more of a switch and steam gamer, but the benefit of using Cortana on my Xbox One is fabulous. I've loved it for a long time. And the funny thing about it is of course also that it's the simplest tasks that are the biggest benefit. Just being able to bark out "Xbox pause" or "Cortana pause" I think it was worth the price for us, but and I can't speak directly to Cortana roadmap since I'm not on the team, but I can say sort of my observations as a fellow customer.

Cheryl Platz: [00:57:51] When I was on the team you know back in the middle part of the decade, there was all this optimism especially around the multimodal phone to PC story. At the time we were just working on what you now see in Windows 10, the Cortana voice assistant, but Windows Phone as we all know has died as a platform and it really changes the fundamental model for Cortana. I think that Microsoft in general is still figuring out what that next step is. I think if you noticed that the sort of shape of the affordance of the Cortana menus in Windows 10, it's essentially a mobile screen on your desktop. Now that the mobile phone is no longer part of this continuum it's an opportunity I think, I suspect, for us to question some of the architectural decisions that were made. Is there a better way to integrate Cortana more seamlessly into this experience? Just as a speech nerd, this is interesting to me the change from "Hey Cortana" to "Cortana" because the word Cortana is three syllables, it's reasonably unique. So I always wondered why you had to have "hey" in front of it. I feel like Cortana was good enough as a wake work. That was very clever that they had chosen a character that had reasonably acoustically unique name. So I'm just glad to see them shortening that because it seems like such a minor thing dropping one syllable, but it makes a big difference. My family actually changed it back to Xbox as the wake word on our Xbox One because they "hey Cortana" felt too unwieldy so I will give it a change back when this change goes live.

Mark Webster: [00:59:33] Back to the comment about kids before, my 4 year old would be super excited to say Cortana instead of Alexa with a hard L. You've been Alexa for at least a year and a half now. So I think you know as a location name it's a great name. And I think you know I am bullish on Cortana as a platform. You know I think there feels like a lot of catching up to do here. But I think you know the cognitive services that power Cortana are you know top in the industry. I have been a cord cutter for almost 10 years, I haven't had cable in 10 years, and my Xbox is my primary media center and I've never used Cortana because I don't have a Kinect, and I don't put on the headset to talk to it. You know and it feels like one minor difference of like having put a mike in the front of it could have changed everything in the way that Cortana could have been the voice controlled media center.

Mark Webster: [01:00:26] You know back to the ad with John Legend, that could've easily been him talking to an Xbox to bring up a show to watch. The Cortana, the shift over I think is the right shift over. I think it's better from a brand perspective. If I was at Microsoft, I would have a lot of concerns about the stuff that's going on with Amazon and Cortana. You know as a user I can't wait to be able to use my Alexa device to control my Xbox to Cortana. But you know there is a bad relationship that Amazon has had with plenty of partners over the last 20 years, and so things like ceding the home to Amazon to TEFs, Alexa coming into Windows 10 machines. You know that's something I would definitely worry me, but I think overall I'm pretty bullish that there are going to be a lot of opportunity for that platform and I think changing the name was a great start to it.

Cheryl Platz: [01:01:23] You bring up a great point about the partnership between the two companies. I was really excited as a tech nerd to see that the two companies had come to sort of an agreement because I think it's a great model for Silicon Valley and good for customers. But the announcement that Alexa might be coming to desktops - that seems high risk in both directions. So it will be interesting to see how that plays out. I think having worked on Alexa it's hard to understate the amount of work that goes into making smart home stuff work seamlessly. So I feel like Microsoft's making out pretty well in that arrangement, like look Alexa you guys you handle that set of things and we'll handle the productivity stuff because we know we're good at it. I'm hoping that is at the core of the partnership. Both companies get a little benefit and customers get to do everything they want whenever they want, but that desktop piece is interesting. We know Amazon has tried to sort of create nonets before and it failed. Will they try again? Will they try to sort of exist on Windows? But Microsoft has also invested in having people use windows so if having Alexa on windows differentiates it from the Mac, that might be enough of a win for them to keep going.

Mark Webster: [01:02:35] Yeah I agree on the Smart Home point now, right, but it feels like you know there's 30 to 40, we're talking about how many smart speakers are out there, there are like 30 to 40 million Xboxes out there. Like there could have definitely been, that whole track could have played out in a different way.

Cheryl Platz: [01:02:54] Oh definitely.

Mark Webster: [01:02:55] Yeah and I also think that Microsoft being under new leadership and changing to a more open ecosystem, like are all the right moves. And so yeah I think it's going to be interesting to see the way all these platforms both compete and play together over time.

Cheryl Platz: [01:03:10] But I will always mourn the loss of the Kinect.

Bradley Metrock: [01:03:14] I better be careful with how I described the Kinect. Yeah the Kinect has had a path that it shouldn't have gone on. Here is what it is. It's just something that was ahead of its time, Right?

Cheryl Platz: [01:03:27] Yes.

Bradley Metrock: [01:03:28] When it came out it was just ahead of its time, and like so many things that are ahead of their time, first it was not used and then second it became an object of ridicule when it should not have been an object of ridicule. And now you know here we are back with voice hardware being acceptable to have in your home and is not being an omnipresent security/privacy violation in your mind and not so the timing would have been much better now, but it's just sort of how things like that play out. Cheryl I will say I can't wait until I can ask Cortana "why am I dying so much in Overwatch? Please make it stop".

Cheryl Platz: [01:04:14] Be careful what you wish for. She'll turn the console off.

Bradley Metrock: [01:04:18] Yeah. Yeah. If I asked her Cortana how I should improve my gameplay in Overwatch and if it just shuts the console off that may not be a message I'm ready to hear.

Cheryl Platz: [01:04:30] But if we talk about personality and these assistants and you look at a device like the Xbox where we know it's not like critical path productivity, it's sort of some interesting opportunities light up and maybe Microsoft could play. This is an interesting scenario we're discussing. If someone engages Cortana in the middle of a game and she's perceived as a game personality, is there room for her to take over your control for a little bit? You could play with that if you wanted to do a fun little design for it, like how something like that would go. It's a more lighthearted note than the death of groundbreaking hardware.

Mark Webster: [01:05:07] It's just crazy to look at you know that the lesson from the Kinect also is that they were all mourning was like the death of a microphone, right. It was like this intense piece of hardware that had all this gesture control and it turns out that the microphone sitting in it was this wildly valuable piece of that experience, that could probably also have been delivered through a different piece of hardware much more effectively, right. So I think it totally was ahead of its time and I think that you know from any product design perspective always thinking about like what are the things we're trying to accomplish? How ambitious is what we're doing here, right? Are there better ways to deliver other things?

Mark Webster: [01:05:48] And I think even Sonos is now going through a lot figuring out how to integrate voice into their interfaces and their hardware, and their hardware has had a mike sitting in there for a while even though it wasn't used, right. So I think from a software perspective there are fewer fingers out here. From a hardware perspective you are definitely not challenged. So think about two years from now it's going to be an even a different voice-first world, and so like what does your hardware does to get ahead of that, because development cycles and product cycles are much longer, right.

Bradley Metrock: [01:06:19] Well I think yes sure and I think there's parallels from the Kinect back to our discussion about the Google marketing campaign. Better tread lightly you know lest you over-promise something that you can't deliver. And the mourning of the Kinect was in my mind the mourning of the promise of what that device was going to bring which was just not ready. And you know and it was society that wasn't really ready for it and the technology wasn't totally ready either. So you know I think Google would be wise to pay attention to that. I see a lot of parallels there.

Bradley Metrock: [01:06:57] We will move on briefly to our last story of the week which is an interesting one being here in Nashville. We know Blake Shelton very well. He is a famous country artist, you might have seen him on The Voice. If you watch that show, Blake Shelton has his own Alexa alarm tones and I just want to ask the two of you quickly and Mark I'll start with you. There is a marketing narrative you know that you could extract from the story. There's a personalization sort of story you could get out of this. What did you think when you read this?

Mark Webster: [01:07:36] The main thing that jumps out at me is just how early this all feels. It's fun and it's whimsical. It feels like remember when cell phone ringtones were like a billion dollar industry and then that changed pretty quickly. It sort of feels like that, it feels like there's going to be some weird moment in time that at some point we look back on and be like remember when country artists were making the voices that you could wake up to the alarm? Snoop Dogg has like a wave's voice and you can get directions from Snoop Dogg. It feels like fun ways to extend your brand as a celebrity to show up in a couple of news cycles. I don't think it fundamentally impacts anything with the way we interact with these things, except for the idea that we move into a world where we have dynamically generated celebrity voices, right, and so especially when you look at Google, Google with Google Assistant is taking this operator model where every time you connect to a third party it's delivered in a different voice. It's not hard to see how all those voices would be celebrity voices. So when I talk to Ford through Google Assistant I hear Mike Rowe. So I think there's like a little hint of what celebrity and voice personality could mean for voice interfaces overall in this news story, but sort of where we are and the idea of like having these custom alarms I think it's just like a fun little moment in time that exists really early and then sort of just goes away.

Cheryl Platz: [01:09:10] I'm glad you brought up the multiple voices thing Mark. It's interesting if you unpack the experience of an alarm, it's one of the least delightful parts of interacting with one of these voice assistants. You are being roused from slumber, probably going to be cranky and you know I think it's really interesting that Google Assistant is using that third party voice model. And I think part of that's to protect their brand and in a way you could kind of look at this as a way to sort of protect Alexa a little bit too. Like why associate Alexa, who we're supposed to welcome into the family with the horrible moment of waking up in the morning? When their celebrities they're perfectly happy to lend their voices to that experience and soften that transition a little bit. But certainly I agree that it's, whimsical is a good word for it. It's not game changing, but you also see Alexa making some interesting plays.

Cheryl Platz: [01:10:02] Going back to the point about the celebrities and the distance from that sort of common consumer, this feels more like reaching the hand across to maybe some markets that weren't thinking about Alexa before, who aren't maybe passionate Amazon customers and getting them to think about things in a different way. You see this a little bit with the Echo Look that we released too. Our target, our sort of ideal customer, we weren't not leading with Alexa, we were leading with their passion about fashion and that might mean they're not your typical tech early adopter.

Cheryl Platz: [01:10:37] So I think this sort of thing is a really valuable way to start trying to extend a hand and get folks interested in this technology that might not have considered it before. I thought it was just a Silicon Valley fad or something like that. So I'm glad it exists in that respect because hopefully it'll bring more interesting customers to the table with new requests, new stories. I'm still waiting for a Peek A Chu voice so if somebody will do that, that'll make my life great. But there is that personal expression piece too; a lot of folks ask to customize the voice. TomTom set up that example you mentioned, the Snoop Dogg voices. This is probably as close we'll come for a long time, but the desire for self-expression in consumers is very much there. They expect it. We see it in games all the time, creating your protagonist as opposed to taking something that developer's built in. If this allows a little bit more of that self-expression, hopefully it'll increase engagement and make that morning moment a little bit less dreadful.

Mark Webster: [01:11:40] I love the idea of voice assistant outsourcing the dirty work.

Cheryl Platz: [01:11:46] Except when it's a horrible task and then we'll let Blake Shelton do it.

Mark Webster: [01:11:52] This is going to be a good cop bad cop, like let me go talk to my manager. It's great.

Bradley Metrock: [01:11:59] It's interesting - you know Mark you sort of glossed over like the ringtones, but yeah ringtones are still a large market. And when I look at this story my impression is, the person runs across my mind is, I sort of view this a little bit differently. I think there is about to be huge money for people with any sort of celebrity status or people whose voice is somehow interesting or desirable. Like I think about the Micro Machine's guy you know, do you remember that? We're all old enough to remember that.

Cheryl Platz: [01:12:35] Oh Yes.

Mark Webster: [01:12:37] Absolutely.

Bradley Metrock: [01:12:37] So you know people with different voices, you know if I could pay, I always think like this, if I could pay you know 99 cents for a ringtone, sure, yeah I'll pay that, song I like or something interesting, I'll pay that, that's fine. The market's sort of established. What would I pay for, you know I like Blake Shelton just fine, but I wouldn't do this, but let's say that I am a Blake Shelton super fan and I want his voice to take over Alexa, not just Alexa alarm tones. You know maybe I might pay you know 99 cents for that maybe or if it's somebody I like or something just for the alarm tones. But what about taking over the whole freaking thing and the voice is now different? You know I could have a Blake Shelton skin for my Alexa, you know and it's Blake Shelton's voice for everything. And with some comedic nuances in there or some different you know personality injected. What would I pay for that? I don't know a couple of bucks and you know what an easy way for celebrities to grab even more money, and also an easy way for IP attorneys, that's another thing that ran across my mind. Think about the policing that's going to have to go into policing the Internet on people's voices being used in appropriate fashions versus inappropriate and unauthorized ways. It's a brave new world. That's what I got out of this, which is sort of taking it a step further.

Mark Webster: [01:14:10] I mean I think there are interesting things we'll continue to see come out around celebrity on that stuff, but I actually like something I could rant about for hours. I actually think the personification of voice interfaces starts to go away. So there there's a design principle called skeuomorphism where you make some digital interface look like the real world equivalent of it. When we all first got the iPhone you know the YouTube app looked like a little TV. The notes looked like a little yellow legal pad and then that starts to go away. I think the idea of like us interacting with these things as people goes away over time. I think this is like an early medium and how we're all getting used to it. You know when we talk about saying like we say "thank you" to Alexa. I think there's a pretty close point in time where we look back at how ridiculous that is. So I actually think that overall we're going to, you know there'll be a place for personification, whether celebrity or not. But I think the majority of your voice you're going to experience will not feel like talking to a person one way or the other. And to go back to the accessibility side of things, it's interesting when you look at somebody who uses a speech reader on like a desktop, they'll listen to it like hundreds of words a minute in a way that feels like a blur to somebody who doesn't do that every day. And I think voice interfaces become much more efficient versus much more personified.

Cheryl Platz: [01:15:36] Yeah I think it's interesting. I've done some voice acting in my free time, and thinking about what you just talked about moving from Blake Shelton doing alarms to Blake Shelton doing the entirety of the Alexa experience, pretty much almost all of Alexa's speeches is dynamically generated in real time. I started thinking down that path and, Blake would have to record enough that they can do synthetic generation of his voice. And then you suddenly you've given up a piece of your identity to a company because their voices are part of us. There's not enough time to talk and unpack a batch thought in this discussion, but I think we will find that there's this tension. How much are we willing to give up as contributors, as creative contributors, Blake or a voice actor, to a company to control synthetically and potentially any arbitrary output they want to do in your voice they can.

Cheryl Platz: [01:16:30] There are some ethical quandaries there and there's some other news articles about synthesizing famous people's voices. Yeah it gets tricky because in many cases we hear someone speak and we believe it is their truth because it's in their voice. So there's a lot there. I don't know as I'd want to see it replacing Alexa's voice with celebrities because I feel like that could be manipulated very easily. But I agree that not all voice assistants will be heavily personified just as Mark was saying earlier. There's a time and a place for it and in some cases I think the Google Assistant model makes a lot of sense. Something less personable and we can look for cues from a customer that they're interested in banter, or are interested in connecting on a more human level as opposed to just turning on the lights.

Bradley Metrock: [01:17:22] A lot of interesting things to think about. Yeah a great commentary and you know I can see it working. Like a ringtone, you get tired of it. You know that for me, after a short amount of time you're on to the next one. Yeah I think personification would be very opportunistic, very situational. You know I would think if you're listening to Blake Shelton's voice, or anybody's voice, the whole time you'd probably get tired of that. It would probably have a negative effect after a while, but yeah it'll be interesting to watch.

Bradley Metrock: [01:17:53] I appreciate y'all entertaining that story here at the end. Cheryl and Mark great to have you on the show. Thank you very much for being so generous with your time and your insight.

Cheryl Platz: [01:18:04] Thank you very much for having us. This was just a very enjoyable conversation thank you Mark and thank you for bringing us together Bradley.

Mark Webster: [01:18:10] Yeah thank you this was a lot of fun.

Bradley Metrock: [01:18:13] For This Week In Voice Season 2, Episode 7 thank you for listening and until next time.