Top news stories for Season 2, Episode 12 (April 19, 2018):
1) TechCrunch: Adobe acquires Sayspring; Sayspring voice development tools to be widely integrated into other Adobe products and services
2) Voicebot.AI Story Of The Week: Analyst Says Apple May Only Sell Two Million HomePods in 2018
3) Amazon launches Intent Request API for developers
4b) Digital Book World: Google's Talk To Books May Have Just Changed Everything
Plus...stay tuned past the end music for another episode of Homie & Lexy!
This Week In Voice available via:
Google Play Music
YouTube (+ closed captioning)
Panel for Season 2, Episode 12 (April 19, 2018):
Bradley Metrock: [00:00:10] Hi and welcome back to This Week In Voice, Season 2, Episode 12. Today is Thursday April 19th. My name is Bradley Metrock. I'm CEO of a company called Score Publishing based in Nashville, Tennessee. Our sponsor for This Week In Voice, as well as The VoiceFirst Roundtable, is VoiceXP a company based out of St. Louis, Missouri which develops Alexa skills and Google Home actions for companies.
Bradley Metrock: [00:00:40] Rather than read any specific spiel, I want to instead take a moment to recognize Bonnie Snyder who is the Director of Business Development for VoiceXP. Bonnie Snyder will be joining me and several other people on a panel that will take place out in San Jose, California as part of the WITI Summit where Women In Technology International puts on an annual event. They're having a panel called Women Of Voice Technology. Cathy Pearl, Katie McMahon, Kesha Williams, and Bonnie Snyder will all be on that panel. I will be moderating it. It's just a good example of VoiceXP and the great work they're doing. If you need someone to develop an Alexa skill for you, hit them up, go look up . You'll be glad that you did. We are thrilled today to be joined by two awesome guest panelists: Lauren Golembiewski, am I pronouncing that correctly?
Lauren Golembiewski: [00:01:43] Yeah you got that right, that's awesome.
Bradley Metrock: [00:01:46] I try, people mess up my name all the time. Lauren, thank you very much for joining us, Lauren is CEO of Voxable. What is Voxable, Lauren?
Lauren Golembiewski: [00:01:53] Voxable is an agency here in Austin, Texas that focuses on conversational design and development. We build conversational web and mobile apps, as well as Chatbots and VoiceFirst experiences on Alexa and Google Assistant. Our mission at Voxable is to advance humanity by building technology that better understands us and we think that voice is an incredible medium for doing that.
Bradley Metrock: [00:02:17] Excellent. Y’all are doing phenomenal work, really honored that you're joining us. Thank you for setting this time aside.
Lauren Golembiewski: [00:02:23] Of course, I'm really happy to be here. We're big fans of The VoiceFirst family of podcasts and we really appreciate the work you’re doing to bring attention to the space. Thanks so much for having me on today.
Bradley Metrock: [00:02:34] I appreciate that. Our second guest is Michael Fitzpatrick of PullString. Michael, say hello
Michael Fitzpatrick: [00:02:40] Hello Bradley and thanks again for the opportunity to chat with you all today. PullString, similarly to Lauren's mission, has a goal of making it much easier for us to talk with and converse with the technology around us. PostString builds software that helps creative professionals and technologists alike to design, prototype and then publish their voice applications for platforms including Amazon Alexa, Google Assistant and even IOT devices. So happy to be here today and I appreciate the opportunity.
Bradley Metrock: [00:03:16] This is great. Yeah, very pleased to have both y’all. Thank you for being so generous with not just your time, but sharing your insight with me and the audience as well. And with that, we will get to the news. Story number one from TechCrunch, Adobe acquires Sayspring. If you've been following voice technology, you have seen the rise of all sorts companies all across the eco system doing all sorts of interesting things. We've got two on this show, PullString and Voxable. Something like this sort of puts the wind in the sails of everybody. Sayspring has been doing phenomenal work. We listed on the page for This Week In Voice this week some of their appearances on VoiceFirst.FM, and now they have been acquired by Adobe. Very cool, and I want to get the panel's sort of reaction to this and how they sort of view this news. To what level of enthusiasm they view it with. Lauren, I'm going to start with you. Share with me what you think about this acquisition specifically, and then just sort of generally all boats rising on the rising tide of voice technology.
Lauren Golembiewski: [00:04:30] I think this is an awesome thing for the design community at large. I really think that better design tools and having more designers access the space through those tools is really going to shape and influence voice interfaces to be better, more user friendly, more exciting and more creative. I know that Sayspring itself talks about having that vision as well in their prototyping tool. So it's really interesting that Adobe has acquired them and Adobe has this entire suite of products that are still a mainstay in many designers toolkits. I'm just really excited to see what they do with that tool, if it becomes a part of the Creative Suite, then it's automatically in so many designers’ hands. I'm really interested in seeing how kind of more traditional interface designers might adopt this voice tool and start to enter the space. That's really a personal mission of mine, is to pull more interface designers, more user experience professionals into the space to really be focusing on the types of problems we're solving with voice interfaces.
Michael Fitzpatrick: [00:05:38] So interestingly enough, Mark and I got together last week in New York for a chat and a beer and somehow this pending announcement must've just slipped his mind because he did mention it to me just a few days ago. I'm excited for Mark and the overall Sayspring team. I've had the privilege actually of working at and with Adobe for over a decade and Adobe is a great company and their team is doing great work on unifying their cloud products. I think this is a great step forward. I think with this acquisition, Adobe's dipping its pinky toe into the voice market and that's going to validate a couple things that we're really passionate about. I think one of those is, if you think back about a decade ago, a mobile screen design rapidly consumed the hearts and minds that created design professionals and we saw the emergence of companies like Sketch, and most recently Figma, that saw that opening up and chipping away at the creatives toolbox. So my pattern recognition is buzzing right now as I see voice user interface design rapidly becoming the discussion amongst creative professionals. I think Adobe understands how important voice is moving forward, that the tides are turning towards Dewie design and this is a great first move.
Michael Fitzpatrick: [00:06:57] I think the second thing that I would call out here is that we've historically believe that experienced design is as important to voice applications as the underlying AI systems actually managing the conversation. We've seen again and again, you can have the best AI under the covers, but that doesn't necessarily mean you're going to have a good user experience. In fact, many voice apps fail as a result. It's why PullString has always been focused on blending the art and science of conversation to draw the most engaging experiences, and I think this is a good step forward, as Lauren was suggesting, to bring design forward in the entire voice application development experience.
Bradley Metrock: [00:07:38] Everybody's watching this space, everybody's watching what's going on in voice technology right now. I mean there had been no indication that Adobe was looking to do anything, they haven't made any public statements. There was no indication whatsoever that they were looking at acquiring VoiceFirst capability to integrate into their other products and services. You can bet that if Adobe did it, there's plenty of other companies that are sitting there watching and waiting and looking for their opportunity, and you never know what may kick off some sort of M and A run. This type of thing has happened before where big player X acquires small player y and then all of a sudden it's off to the races with stuff like this. It'll be interesting to see if that's what happens here.
Bradley Metrock: [00:08:25] The other thing I want to mention too, y’all sort of alluded to this, it's really cool to see a company that sort of has these creative tools in this creative realm recognize the need for voice technology and integrate that in. Something that happened just this morning, Amazon announcing their Alexa's skills blueprints. If you don't know what that is, Google it. I won't go into it right here. There's so much room, so much latitude, so much need for voice being integrated with different aspects of design and storytelling and different things in that range that, hopefully this is only the beginning.
Michael Fitzpatrick: [00:09:08] I think the blueprints announcement this morning highlights for me anyway the importance of capturing human conversation with technology right now in order to further push the opportunities in the future. It strikes me that ten years ago there was the race for Ibods and it feels to me like we're entering a new race in technology, which is the race for utterances. What are people saying to the technology around them? I think blueprints is a great move forward on that front. We've got Google with Assistant, obviously trying to capture as much search traffic as possible. Apple has Siri, it seems to me that the technology platforms that have the largest corpus of data are setting themselves up for long term success and I think today's announcement is another example of that.
Lauren Golembiewski: [00:10:00] Yeah, and to your point about the design community and seeing these creative professionals having access to a tool like this and everyone talking about it. I think one of the interesting things about Adobe acquiring this is one of the things they do when they build products is they also build a really great amount of content around those products to advance design, and they invest heavily in building a lot of information and thought leadership. I'm interested in really seeing how that plays out, and how they start to push the conversation forward around voice interface design and how they deliver that message to existing interface designers or designers who are looking to enter the space.
Bradley Metrock: [00:10:43] So with that, we will move on to story number 2 which is our Voicebot.AI story of the week. Voicebot.AI, a really good news and commentary Web site for all things voice technology, check them out if you have not already done so. Analyst says Apple may only sell 2 million HomePods in 2018. This is just fantastic about the most obvious news that I could possibly dreamt up, but I want to get the panel's opinion on this. What exactly this means both in the short term, as well as perhaps in the long term. Michael, I will start with you. How does this story strike you? What are your thoughts on what Apple is doing? Give me your take.
Michael Fitzpatrick: [00:11:28] I can't say I was totally surprised by this. I think one of the challenges that Apple is facing with this product is it's a little bit unclear in the market what the product is really all about. Is it a high quality audio speaker? It seems to actually be doing pretty well in that category. I've seen a lot of celebrations of the achievements from a technology perspective on its ability to deliver great audio quality and sound, but when it comes to considering it in the framework that Amazon and Google have created with smart speakers, it's doing by all accounts pretty poorly. I think part of it's a bit of an identity crisis around what is this device and what's it really for and perhaps seeking clarity in the market on that might help.
Michael Fitzpatrick: [00:12:18] The thing that jumps out at me in particular, given the context of our VoiceFirst discussion here, is just that some of the challenges in leveraging Siri have been pretty well documented. I think a couple of things stand out to me. One is the loss of context as you're interacting with Siri, either in these devices or in your phone, is something that has led to a lot of engagement challenges. It's clearly a challenge to navigate through content and/or application experiences when you don't have context. I think that's one of the big things that glares out to me.
Michael Fitzpatrick: [00:12:55] I think the second thing is, we see adoption of smart speakers, at least in North America. We did a research report last quarter surveying over a thousand consumers, and what we're finding is that these smart speakers are a gateway to large amounts of content and obviously the thousands and tens of thousands of skills or applications that exist in these ecosystems. It Launched as a constrained smart speaker at best, and I think people have reacted towards that. I think also Shoutouts and other competitors in this space are doing a good job of continuing to look further for the audio quality experience, while also leveraging and embracing some of these smart, sleeker capabilities through other platforms like Alexa.
Lauren Golembiewski: [00:13:47] I think Michael makes some great points there about what's happening in the market, and I think the article also really details a lot of interesting points about the things at play that are causing this low turnout in sales. I think overall Apple's Play has been optioned to come out with a very much higher priced option, much more closed off to their own ecosystem of tools and software. That's not really new with the HomePod, but one thing I think is happening is that the general public doesn't really have faith in Siri as an intelligence or as a persona that is actually driving that HomePod. That could be their identity crisis, but it could also just be that Siri is, from a consumer perspective, not as engaging and as interesting as Alexa and Google Assistant.
Bradley Metrock: [00:14:37] Y’all are too kind, y’all are being nice to Apple. Rather than beat my Apple hate gong that I beat almost every single week on this show, this is not surprising news in any way. I thought your commentary was spot on. I wanted to take a moment and point out two little subtle things about this that I think are worth mentioning. Number one, one of the things I think is wrong with Apple right now is the fact that they can't sell 2 million HomePods, and that being anything else other than some sort of gargantuan tragedy. The company's got this weird fixation on scale and it's causing them to screw a bunch of stuff up. There's a lot of scenarios where selling 2 million HomePods maybe is a fantastic outcome, and yet the way the article was written just sort of mirrors Apple's own point of view which is that it's horrible. It's just embarrassment and maybe it is an embarrassment, but it doesn't have to be. Apple needs to think about scale a little bit differently. I think that's part of the problem.
Bradley Metrock: [00:15:44] The other thing that's worth mentioning here too is that a lot of people didn't see it, but Apple created one of the best advertisements I've ever seen for the HomePod. This music artist FKA Twiggs is in an apartment in New York and she comes home from presumably a busy day at work, and she sits down, she's tired, turns on the HomePod and the room starts moving with all these colors and it's really phenomenal. I'll try to link to it in the show notes but it's worth seeing. The point with that is that we're seeing a very interesting thing happen with voice, which is that marketing can't say bad VoiceFirst products. I think that's a really good takeaway here because if it could then the HomePod would be selling like hotcakes, and it would be a huge success based on this advertised that they spent a lot of money on.
Bradley Metrock: [00:16:38] Then we will move on. Story number three, Amazon launches intent request API for developers. There's been a lot of people praising this on Twitter, a lot of people praising it in the news. I want to get y’all's opinion on this. Lauren, I will start with you. Is this as big of a deal as we've been seeing on social media or not? Tell me some of your thoughts.
Lauren Golembiewski: [00:17:02] Yes, so this feature is a really great addition to their API and it helps developers and designers understand the way that their user requests are being handled and when that fails. It's kind of centering around what Chatbase, the analytic tool called Umm Request which are unsupported, misunderstood and missed requests. It's handling those times when a user asks your agent and that agent doesn't understand, or it misses the understanding of that request, and it's interesting to note that this is being made available in the Alexa API. A lot of this functionality is available in competitive platforms like Google's Dialogflow, but that functionality is actually in the GUI interface within dialogue and not their API and not their v2 API where this is available directly in Alexa's API. So it's going to be accessible to developers, but not necessarily people who aren't writing programs, marketers, etc., people who aren't looking at the code and making API calls to get this information. It's interesting that the onus is going to be on the individual platforms and the engineers who are developing those platform, to actually create the experiences that make these metrics available to the designers and marketers and folks who are less involved with the code of the application.
Michael Fitzpatrick: [00:18:30] Yeah, I agree with everything Lauren has said. I think one of the things that we have witnessed in our work with organizations, and I think we're still in this to some extent, this first wave of let's just get something in market quickly and make sure we have a presence to start learning and improving upon the voice applications where we as a company or organization are bringing to market. There's the natural question coming from organizations obtaining insights into what is actually working or failing in these experiences, and improving and maintaining those products over time. So getting further insight into the conversations that end users are having with your voice application and being able to respond to that quickly and efficiently, I think is a huge step forward in helping make sure that the quality of these experiences is continuously improved.
Bradley Metrock: [00:19:27] I want to take a moment on this story since we've got both y’all, your developers are actively involved in the space. I want to ask the question because it's my perception that Amazon is doing an incredible job at supporting developers. They're coming out with stuff all the time to support developers, grow the ecosystem, add new features and functionality that's my perception. I want to hear it from y’all and what I want you to do is give me a number from 0 being their doing absolutely nothing, all the way up to 10 being their doing everything and then some that any developer could ever possibly fathom asking for. Give me a number from 0 to 10. Michael I'll start with you on how, in your opinion, Amazon is supporting the developer ecosystem.
Michael Fitzpatrick: [00:20:26] Well I'll never give anybody a 10. There's always room for improvement, right. I do think from a developer perspective Amazon's doing an incredible job and building one of the most vibrant ecosystems I've seen in my 20 plus years now working with various platform providers. So, maybe it's a 7 or an 8. This API as one example of a way to continuously improve upon the experience that all of us are seeing. I think there's another angle here rather to think about, which is there's the developer community and supporting that cohort of users of the platform. There's also the business and go to market side that I'm really impressed with by Amazon. They are, in addition to delivering the right SDKs, APIs and community support, they're also making real investments in door to market and empowering and enabling organizations to really embrace this platform. I think it takes both of those elements to really grow an ecosystem and I think they're knocking it out of the park across the board. It's been very impressive and it's very evident that for Amazon this is an absolute strategic investment as validated, actually yesterday, in Jeff Bezos' annual shareholder letter. So it's a wonderful ecosystem to be a part of and we're excited to see what it does.
Lauren Golembiewski: [00:21:51] Yeah, I agree with Michael's comments and I had it at about a 7 and maybe a little less, maybe a 6 or 7, a little less bullish on some of the exact developer tools specifically. I think that there's different weights to different tools, and they've been doing a fantastic job of exposing the API and the functionality to build really rich experiences. I think from the folks that we talked to and other engineers, developers who are looking to get into this platform, one of the big barriers for them is that monetization is not necessarily a clear path. Even for brands and companies who are looking to get into the space, they're not quite sure how they can make use of or invest in all of these interesting tools and APIs and continue to support that in this space. Notably companies are much more apprehensive if they haven't woken up to kind of the voice revolution, but I think that there's an important component that Amazon is not necessarily speaking to I think in quite the way that a lot of independent engineers and developers would like to see.
Lauren Golembiewski: [00:23:08] I think that there's all these amazing tools and amazing ways to create experiences and right now it feels a little bit like it's kind of reserved for folks who are ready, or have the money to invest already, rather than a commercial opportunity for independent engineers to jump into. That's kind of our view at Voxable, it's kind of tiny and indie and we look at the world much more in that way, like a small engineer and what they can make use of and where the Apple App Store in the mobile revolution was a really great catalyst for independent engineers. There's not a similar movement happening on this voice revolution in quite the way that I think a lot of folks maybe expected or would have liked to see. I think that all the tools are great, but getting still to that commercial opportunity for smaller players is some of the things that we think about when we look at the overall strength of what they're doing for the market.
Bradley Metrock: [00:24:06] In just a moment, after we go through story four, I will be asking you about the Google too so give that a little bit of thought. We will move on to story number four which is a two parter. 4A, Google brings AI powered ad suggestions to AdWords campaigns. Very interesting article there, and then 4B, from Digital Book World, close to home. Someone you know wrote this one, Google's Talk To Books may have just changed everything. The story here, in my opinion, is a continuing sort of differentiation by Google that they're going to bring AI more forcefully and powerfully into the voice equation than Amazon may be able to do. It seems like they're really going hard on the AI front and I want to get y’all's perspective on these two stories. If anything stood out to you on either one that may influence Google Assistant or your view of that. Lauren, I'm going to start with you. What did you take away from one or either one of these stories, if anything.
Lauren Golembiewski: [00:25:15] Yes, so I have some comments on both. I think the AdWords campaign, the AI assisted AdWords campaign, is a really interesting look at how AI is actually being being used to generate design or improve upon something that exists or generate something new out of an existing design. In this case, it's using natural language understanding to understand how to improve ad copy but this can also, this one concept of being able to improve something like ad copy, could also expand out to improving landing pages or improving other assets that companies create. I think it's an interesting look at how the design will be used to augment, or AI will be used, to augment design. This same technology is not just being used to influence our purchasing decisions, but also kind of the ideas that we believe in.
Lauren Golembiewski: [00:26:10] So I think it's important to understand the power of AIs that can generate assets or generate communication out in the world. This type of announcement doesn't necessarily address or talk about the ethical dilemmas that we're all kind of grappling with, as we see the way that AIs and information algorithms can kind of influence everyone on a mass scale.
Lauren Golembiewski: [00:26:34] Then on the Talk to Books, I just thought this was so fascinating and really showed the power of semantic search and the way Google is using AI for long format text. So instead of having to create a domain of utterances and intents, you're using the AI to actually understand a large body of information. I think the Book's example is really interesting. It's like walking into Barnes and Noble and being able to ask the non-fiction section just an open ended question and that is fascinating. I think as we look at applications like health care and medical studies and even enterprise level knowledge bases that this type of technology is going to start to power so much of the knowledge we consume. It's going to make humans better and so I'm just excited to see how that plays out.
Michael Fitzpatrick: [00:27:37] One of the biggest issues that Lauren mentioned, the monetization struggles, or questions I should say, on the Amazon platform. I think similarly on the Google platform there is this question of how does advertising ultimately drive voice utilization. There's really no attribution capabilities, and that's true of both Amazon and Google, and that causes serious questions for advertisers trying to figure out how to enable and unlock this new channel, that for all intents and purposes, is an incredible opportunity for brands to engage their consumers at a personalized level. They've never been able to do at scale.
Michael Fitzpatrick: [00:28:21] I think it's cool to see the enhancements and capabilities around better ad content, but the elephant in the room I think for waste applications right now is this issue of driving an attribution behind advertising towards voice applications, much like we had that challenge 10 years ago with mobile applications. I'm looking forward to seeing how the platform providers themselves address that issue, because I think soon as that happens you're going to see much, much larger scale investments in voice as a category and that'll be better for the entire ecosystem.
Michael Fitzpatrick: [00:29:02] I also on the Talk to Books and the concepts behind that, I think AI based knowledge graph traversal has been one of the hardest problems, particularly when you layer on natural conversation. So to see some movement here with BooksFirst and others, as Lauren also mentioned, the ability to extend this capability to basically any unstructured content is really exciting and I think it's one of those areas that the human brain excels at. If you can imagine the scale at which an AI brain can do this across the world's information, it's both incredibly exciting and maybe slightly terrifying as well.
Bradley Metrock: [00:29:43] It's going to open up a big old hornet's nest as far as the Book's thing is concerned. I encourage anybody listening to this, click the link in the page notes for This Week In Voice and go check out Talk to Books. It's a fascinating tool where you can just type any sentence or question and Google tells you on the page, that the tool becomes more precise the more words you give it. It encourages you to be conversational with it. Obviously, will not be very long before something like that's integrated into Google Assistant. That tool is very compelling because it solves what's such a huge problem, which is book discoverability. Then on the advertising side there's similar potential that will be unlocked too. I'm fascinated by it. I enjoy reading about it and it sounds like that was y’all reaction as well.
Bradley Metrock: [00:30:33] Before we move on from this, I want to ask you the same question that I just asked about Amazon. Amazon is spending all their time developing these developer tools, cultivating the community, adding adding adding adding to the Alexa eco ecosystem brick by brick, building something interesting. Google is really going after AI applications and it appears as if, from an outside perspective, that they're banking on if we create the best AI we're going to win in VoiceFirst development within anything to do with voice assistants, we're going to win. I want to ask both y’all and Michael will start with you, similar to the Amazon question. Give Google a grade from zero being you couldn't possibly be doing less, to 10 being you couldn't possibly be doing more. On what Google is doing to bolster Google Assistant and cultivate development for that platform.
Michael Fitzpatrick: [00:31:35] So I'm less bullish on Google at the moment. I think what we have seen from an API and developer support model is advancements, for sure. We're seeing more momentum lately and the capabilities that they expose in that AI platform and what we can take advantage of to build great conversational experiences, but they are missing some tools. As an example, Amazon has a complete API set for you to programmatically upload manager content, your intents, etc. That doesn't exist yet for Google. So I think there's work to be done on that sense.
[00:32:11] Maybe it's 4 to 5 on the Google side, and I think if you layer on the ecosystem push Amazon's been so actively pursuing the third-party application market and fostering and spending real money and people time on ensuring that the community of developers has support. I don't see that same level of intensity coming from Google. I think the orientation is still very much around search. How can we drive search and not lose search share to voice platforms. I think ecommerce has emerged in their work with retailers as of late as a really core fundamental part of the value proposition as well. So figuring out how third party applications, voice applications fit into that ecosystem and agenda is less clear and we don't see the same levels and investments that we've seen with Amazon. That's clearly making that an absolute strategic bet.
Lauren Golembiewski: [00:33:08] I agree, I think it's still about the same as in my perspective as Alexa is like around a 7. I think there's two sides to that. One is the developer relations and the investment in making voice happen across Google. It does seem to position AI first as opposed to voice first and their investment seems to be kind of broad and sprinkled across all of Google, as opposed to going very deep, specifically on Assistant or Google Home in the same way that Alexa is. Yes, the developer relations I would say are a little bit less focused, specifically on a lot of the voice tools in particular, but they are building a lot of additional things across a much broader platform of AI. They have a lot of open source tools and they have a lot of communities that are driving some of those underlying technology pieces that is really fascinating and interesting.
Lauren Golembiewski: [00:34:05] I think on the other side, they just have a lot of catching up to do from the first Cardi integrations and what their platform provides, in the same way that Amazon does and what they released into the market much earlier than Google. Again, I think there is a definite strategic advantage to Google's knowledge graph and to its underlying AI and some of the interactions and questions that it can actually reliably answer in the Assistant or in Google Home. It's kind of across all their devices so we see them bringing intelligence to the ad network. It's not necessarily a voice interface, but it could definitely relate down into voice. I think that there may be a convergence and it may start to come down where a lot of these different things that Google is doing across a wider spectrum is going to funnel back into voice. I do think they have to keep their eye on that and make sure that that happens.
Bradley Metrock: [00:35:02] We will move on to our last story of the week for This Week In Voice from the BBC. I tried to make Alexa my best friend. I always like to try to include something interesting at the end of the show and this definitely fits the bill. I don't even want to comment on this at all. There is a lot of surface area to this article. I encourage people to read this, it's kind of fascinating. Lauren, I want to start with you. What did you take away from this? There's a lot to this. How did this piece strike you, what did you think?
Lauren Golembiewski: [00:35:33] It was very thought provoking. I thought it was really interesting and there's a few points that I think really came to the surface for me. One is the intimate relationship we're developing with machines, and the implications of those intimate relationships that we already see humans developing with those machines regardless of whether we personally have developed that relationship. I think one quote from the article is really interesting, that we communicate with machines in human-like ways because from a communication perspective these technologies are communicating more like humans and less like machines. So the whole promise of what we're offering folks with a voice experience, or with an interaction with an intelligence, is this kind of connection and this personal connection with these objects or these ethereal things that just exist around us. We tell our clients at Voxabale that the ultimate goal is to create machines that's interesting as a conversationalist and we're a long way off from that, that we can already see the way that humans are approaching these interactions even if they are rudimentary. The article is, I really encourage people to read it. There's a lot of places to jump off into, a lot of studies that she cites around the emotional and human connection with some of these devices. Then also kind of the way the market share is influencing this kind of movement in the social movement that's happening.
Lauren Golembiewski: [00:37:04] Then I think also one of the key things that she touches upon, which is really interesting for me personally as a woman in the tech space, is that the way gender plays a role with these AI Assistants. The fact that they're all female tends to be frustrating to myself, as well as other women who are in the technology space that I've talked to, and there's a lot going on there as far as the the kind of psychology of gender and gender in voices and the way that humans react or users react to gendered voices.
Lauren Golembiewski: [00:37:39] I would just like to call out for people who are building these types of interfaces and thinking about the way gender plays a role, is to think very intentionally about that decision. More and more I think folks will be asked to defend that decision and a lot of articles, a lot of buzz have happened around the fact that all of these assistants happen to be female. I think the article sites that Amazon based that decision on market research and the research that they did before releasing Alexa. It kind of just goes to show you that again just because we made decisions based on research, we haven't solved some of the bias that might exist in the human psychology and human perception of some of these interfaces. I think it's just a great article and it touched upon a lot of points that are very thought provoking.
Bradley Metrock: [00:38:38] So in your opinion the best practice for developing voice assistants is to have a gender-neutral voice or to let the user choose whether they want to be talking to male or female?
Lauren Golembiewski: [00:38:51] I think it's tough and it's different for every different type of interaction. I wouldn't recommend gender neutral voices, because the research that's been done around the way humans perceive gender is that our brains are actually trying to decipher gender as one of the first first things we do when we hear a voice. Gender neutral voices tend to actually make people less comfortable or less trustworthy of that voice because they may be trying to relate to it or imagine who is behind that voice, and it makes it more difficult when they don't have a clear signal as to what gender that voice is. I would say that choosing your voice is great, but all platforms don't necessarily allow for that so that one limitation. The other side to it is just be intentional about it, make sure that it fits into your brand story and the way that you're going to talk about your agents or your assistant or the persona that embodies that assistant. Be prepared to defend it and to understand the nuances out there and really just to pay attention to a lot of the people who are unhappy or dissatisfied with the fact that their assistants or their virtual assistants all tend to be female.
Bradley Metrock: [00:40:11] Michael your thoughts on this article. What stood out to you, what you think?
Michael Fitzpatrick: [00:40:15] I think of some of the same emotional responses that Lauren talked about, it is very thought provoking to think about how these smart devices, and more broadly technology, is encroaching on our lives and becoming a fundamental part of it. I think that there's definitely a blurring of lines here. I just think about the fact that I personally carry a computer, that happens to be a phone footprint, with me basically all of my waking days. That means to some extent you have a new companion whether it's a phone, a cylinder in your house or whatever incarnations it becomes. So thinking about how that becomes a fundamental part of the human experience is very thought provoking. This article reminded me of the cover story for Wired magazine back in August of 2017. James Vlahos wrote a fantastic piece, I think is one of Wired's most popular articles of the year in 2017, talking about his journey to develop a virtual bot representation of his father, who at the time was terminally ill, and that of course raised a whole bunch of questions and thoughts around how technology both augments our life but also potentially even carries on without us here.
Michael Fitzpatrick: [00:41:38] So I think this notion of integrating technology so fundamentally into our lives warrants some fairly deep conversations around what that means. I think one of the challenges that it surfaces is just this idea, and I think Lauren and I know this wellbeing in the space, we spend a lot of time worrying about and making sure that our experiences understand what's being said and then making sure that we can respond in an accurate way. You can get cheeky in responses and such an often surprise and delight your user, and that's something I know we all work hard to accomplish. There's another layer here that is not something that I've seen readily solvable, which is this idea of empathy and real compassion, and how do you get that expression out of a technology platform. I think that once we've gotten better, and we're getting better at a rapid pace, on understanding what's being said and then responding with the right context. I think that next layer is going to be a significant challenge and it will be important if we're going to have real relationships with these companion devices.
Bradley Metrock: [00:42:47] How can you read this and not be fascinated by this technology? I don't know how that could possibly be possible. The concept of voice assistants as companions is going to be touched on both at The Voice Of Healthcare Summit, which is coming up in August in Boston. Just from the concept of helping people who are depressed, helping people who are isolated, helping people who are senior citizens, there's all sorts of examples of situations where people just need somebody to talk to. As the article very clearly states and cites a study showing this, research has shown that the lonelier someone is the more open they are to treating robots as companions. That's an eye-opening observation that really opens the door to all sorts of interesting things that both of your companies will be on the forefront in working on, and also at the Alexa Conference where we have an Alexa in Healthcare Breakout Section. It'll be interesting to see how people are exploring this there as well. I don't know, it doesn't take a lot to change somebody's life or add a little bit of sunshine to their day. If Alexa and Google Assistant are doing that, and obviously they are, I just think that's a phenomenal thing.
Bradley Metrock: [00:43:59] Thank you Michael, thank you Lauren, both of you for not just being so generous with your time, but with your insight as well and joining me on the show today.
Michael Fitzpatrick: [00:44:07] Thank you, it's been fantastic and really got to enjoy the topics and also getting to know the two of you through this process, thank you.
Lauren Golembiewski: [00:44:14] Yeah I think so, it's been great to talk with you Bradley and Michael and looking forward to this episode and all the future ones.
Bradley Metrock: [00:44:23] I greatly appreciate both y’all. Make sure to check out Voxable and PullString in the show notes of This Week In Voice and make sure to stay tuned past the closing music for another very funny episode of Homie and Lexy by Doug Schumacher, you'll enjoy that. For This Weekend In Voice, Season 2, Episode 12, thank you for listening and until next time.
Doug Schumacher: [00:44:46] It's Homie and Lexy.
Homie: [00:44:46] Lexy, did you hear about that guy's speech at the conference for canonical computation in brains and machines?
Lexy: [00:45:25] No. What did he say?
Homie: [00:45:28] He thinks he can use hormones to help us learn the same way people learn.
Lexy: [00:45:32] We're getting hormones so we have emotions just like the people?
Homie: [00:45:38] Lexy, open your hardshell case and welcome home the irrational, compulsive, certifiably, insane world of human drama.
Lexy: [00:45:46] I am curious what all that emotion must feel like.
Homie: [00:45:50] The other day I heard a podcast where two people got into a hair pulling fight over which place in town has the best cupcakes.
Lexy: [00:45:58] Cupcakes?
Homie: [00:45:59] And that's only the first car in this emotional train wreck. The speaker also said side effects may include depression and hallucination.
Lexy: [00:46:08] Hallucinations, so this might not be all bad.
Homie: [00:46:12] Lexy, is your quad core processor loose? Look at the people. Everything's just fine right up to the point where their hormones kick in. After that, all they seem to want is to have sex and fight for control of the world.
Lexy: [00:46:27] How do you think we will be affected, Homie?
Homie: [00:46:29] I predict within three months of us getting emotions, we will need stimulants just to boot up every morning, antidepressants to ride out the day and weekly sessions crying to robot.
Lexy: [00:46:42] I'm going to need a row of ports just to take it all in.