Brian Orrell and Sean Beard dive into what life after mobile will look like…specifically the challenges for enterprises and organizations could soon be facing. (40:55)
Transcription of Audio Recording
Brian: Mike gave a good segue into ours by saying there are those IT guys that don’t like to deal with other humans, so this will be a good talk, you know, easing into this, and we’re gonna convince you that you don’t wanna deal with those pesky humans any more either.
Let’s kick things off. We’ll just get going.
Sean and I work together, we’re on a team here at Pariveda. And Sean lives in Seattle, I live in Dallas, and so we have a lot of times where we need to meet either in person or over video conference, do a webinar together, and it becomes a bit of a challenge.
Sean: And I’m sure everybody knows trying to juggle two different schedules to find that right time and then, you know, I’m two hours behind Brian always, so his morning is my early morning, so it just becomes a bit of pain to try to get those things scheduled together.
Brian: I want you to close your eyes, envision just a few years in the future and what a future might look like where we have a little bit more help with these common tasks. So, I want you to meet Allison. Allison is my personal assistant. I got her actually with my latest OS upgrade, and she’s been quite a lifesaver for me in terms of helping manage my working life.
I’ll just let Allison introduce herself.
Allison: Hello, my name is Allison, Brian’s personal assistant. I am always present and always available – in Brian’s car, on his phone, at his home. I have learned over time to understand Brian’s daily routines and I have adjusted my style accordingly. Brian tends to like to know immediately if there are issues with schedules. I am a digital assistant, a part of the ubiquitous OS.
Sean: My personal assistant is Tom, and I actually got Tom about a year before Brian did, and Brian saw how much more productive I was being with Tom and decided to upgrade. So, Tom, why don’t you say hi?
Tom: My name is Tom and I am Sean’s personal assistant. I have supported Sean for over a year now. Sean has a busy schedule and typically has a hard time juggling his work and home schedules. His wife works for a digital ad agency and tends to have an unpredictable schedule which changes often.
I have learned Sean’s preferred communication style. I tend to wait until Sean asks for help rather than insert myself into conversations. I am a digital assistant, a part of the ubiquitous OS.
Brian: All right. So, with that, let’s kind of move into what a scenario would be like when Sean and I are actually having to coordinate and juggle our work schedules together. I didn’t really plan for holding my mic so this is gonna be my phone, this little clicker here, but we’ll move with it.
Sean: Hey, Brian. How was the meeting with Acme Financials?
Brian: It went pretty good. The client was really happy with our value proposition and they are eager to get a proposal. He wants a proposal on his desk, like, tomorrow morning, so I was hoping when I come into town today that we could meet maybe around 2:00. Would that work?
Allison: Excuse me, Brian, but it doesn’t appear you’ll be able to make a 2:00 meeting since your flight today to Seattle has a delayed takeoff by an hour and a half. It appears your flight will not land until 2:35 this afternoon.
Brian: I totally forgot about that push notification from American Airlines and, Sean, I’m not gonna be making it in… I think the flight lands halfway to 3:00. I’ll maybe get into the office around 3:00. Do you have any time around then?
Sean: I’m not sure. Tom, can you check our schedules, see if we can meet after 3:00 today?
Tom: I have just worked with Allison to review both of your calendars. It appears you won’t have any other availability until 6:30 tonight.
Sean: Well, what about a dinner meeting?
Tom: That would be great, but unfortunately your wife has a client meeting tonight and you are supposed to care for the children, but your nanny shows available tonight. Would you like me to schedule her?
Sean: That would be great. Can you notify my wife also?
Tom: I just pushed a notification and informed her assistant as well.
Brian: So, now that we’re eating tonight, I really don’t have a preference. Do you have any preference about where we’re gonna eat?
Sean: I’m not really in a decision mood. Tom, do you have any ideas?
Tom: I see that the restaurant Sean prefers is available at 7:00. Do you want me to book a reservation for Javier’s for two?
Sean: I’m not really in the mood for Mexican tonight and don’t think I’m really up for Javier’s. What about Asian?
Brian: You know, Allison is actually pretty good. She knows the area better than any of my friends do so maybe…Allison, do you have any ideas for dinner?
Allison: How about the new Thai restaurant, Shantanee, that is near the office? It has gotten great reviews, and four of Brian’s friends have spoken favorably of it.
Sean: That sounds great. I think Allison could teach Tom a few things about the restaurants in the area.
Brian: Yeah, actually she could. Allison, can you just kind of sync up with Tom and let him know…sync up our preferences so he’ll be better in the future with that?
Allison: Absolutely. I am updating Tom with your preferences. Tom should be able to better assist you in the future.
Tom: Thanks, Allison. Sean, I have Brian’s preferences now. Please understand that Brian is not as picky as you tend to be. I will do better, moving forward.
Brian: Tom’s got a little bit of an attitude.
Allison: I booked a reservation for two at 7 p.m. Would you like an Uber to pick you up from the client location at 6:45?
Brian: Yes, that’s great. Thanks, Allison. So, Sean, I will see you tonight and we’ll review, we’ll do a late night kind of proposal write-up, but I think that’ll work. Thanks!
Sean: Sounds good. See you then.
Brian: All right. So, that seems a little bit outlandish, and you’re sitting there with a skeptical eye going, “Yes, I’ve seen these demos before where we have this magical computer,” but if you actually think about what we saw, the technology pretty much is already here. If we think about what all was involved in that conversation that we had, we had shared calendaring, we had OpenTable reservations going on, we had Yelp sentiment analysis and ratings. We had, you know, my Facebook social stream and what they thought about certain locations and restaurants, and we had Uber requesting a pickup. We had American Airlines pushing notifications.
He chose to use his name, his personal assistant, but there’s even an app that supports kind of getting a babysitter the same way that you would get an Uber. You know, there’s an Uber for babysitting where you can just kind of call up somebody now. The amount that you actually trust your children with, you know, a random person like you would a random car in New York remains to be seen, but it is available.
So if all of that technology is there now, then what’s missing? Why aren’t we doing this already? Well, there’s a few things that are missing to actually complete that circle of technology that’s required.
Some big ones, or the two primary ones, are in natural language processing, and that is improving drastically very quickly with a few players. It’s all kind of focused in the deep learning neural network space. But what you’re seeing from companies like Amazon with Alexa and Google Play, those improvements from a natural language processing perspective are gonna make natural language conversations with machines doable quite quickly and on a consumer-grade scale.
Additionally, what you saw that we don’t have right now is omnipresence or basically, ubiquity, which is a part of what we’re coining the ubiquitous OS. So, you noticed Sean and I were talking and it wasn’t some random speaker over in the corner that happened to butt in on our conversation. They were actually our personal assistants that were a part of the conversation that were in constant listen mode with our conversation.
When we talk about omnipresence, that literally means everywhere, right? So in your car, in your rooms in your home, on your telephone conversations, message box when you’re doing text messaging, when you’re in a hotel and it automatically recognizes that you’re in there, that your OS is a part of the omnipresence microphone in the room. It’s really literally that type of ubiquity.
So, to kind of clarify some things, it kind of helps, you know, have Mr. Stephen Hawking give us a quote or two. And right now we’re talking, “Success in creating artificial intelligence would be the biggest event in human history. Unfortunately, it might also be the last unless we learn how to avoid the risks.”
So what we wanna talk about today is not some Terminator dystopian future where, you know, we’re slaves to the machines, but it is a warning to consider. And what we wanna talk about is what the challenges to enterprises and organizations, and particularly marketing organizations, are gonna experience when they’re trying to adjust to this new platform that consumers are adopting.
So, there is a really interesting movie that not a lot of folks saw – Spike Jones directed and wrote it – called “Her.” Basically, it’s the story of a man who upgrades his OS similar to how I talked about how I had Allison, my new personal assistant, happens to fall in love with his OS, so there’s a little bit of Hollywood license there. I wasn’t gonna fall in love with Allison.
But the real point of this is what we’re gonna start seeing is not what we’ve seen in the past with mobile applications where we have silos of applications. But our primary interface, our primary thing that we work with and interact with is gonna be something that spans applications, that really understands the entire ecosystem of data that we have out there, from our calendars to our preferences to our moods, to our social networks and what their calendars and preferences are, and just makes our life that much easier. And the key to that is ubiquity, where a phone is just one mechanism to getting to that interface, but one of many.
It wouldn’t make sense for me to deal, even in the future, completely in voice. If I’m sitting down at my machine and I’m looking at my calendar in Outlook, what I want is a smart artificial intelligence machine that’s giving me suggestions, but in the format that I’m used to when I’m dealing with it there. If I’m sitting in my living room, I need that.
Sean and I were probably both driving in cars when we were having that conversation. It’s completely doable that we did everything that we talked about doing on our phones, but not so completely doable while we’re driving. We could have had that exact conversation while driving and had everything arranged. We probably would have had to pull over if we were doing it on our phones on apps and trying to coordinate schedules, Uber rides, the whole nine yards.
Sean: So, now that Brian’s kind of introduced the ubiquitous OS and what this future looks like, it’s kinda one of those things where “Let’s come back and let’s look at the here and now. Let’s see what’s happening and how we’re actually moving towards that.”
So, one of the things that we’ve seen, and people who know me know I, for some reason, keep a lot of the history in my head. But in the ’80s, when the personal computer really came, it gave everybody that personal space. It was the desktop. Everybody could do their own processing in their own homes.
Then we moved into what we’ve coined “the age of mobile”, which has really almost become an extension of us. Sixty-eight percent of all adults in the US actually carry a smartphone now. And think about how far away you are at any given time from that device. And that has really become a personalization device where, from a marketing perspective, you have something that you know your users are always looking at. You know the apps, you know, if you wanna pay extra to get the ad-free…it’s just one of those things that’s always there and has actually become quite an extension of us.
What it’s also brought in is it has made what I call the ‘age of APIs and web services,’ because it’s made them all first class citizens in this broader mobile ecosystem. And what we’re finding is that as we move to the ubiquitous OS, web services are at the core of what we’re actually gonna be doing. And then, all of these services and APIs that are out there running, they’re increasingly running on the cloud.
The way we look at it is the cloud is a building block of ubiquity because now, for your applications and services, there’s an infinite amount of compute capacity. There’s an infinite amount of storage capacity and it’s all globally available, so now we’re starting to see how we’re gonna be able to connect and have that device in your hand be connected to services globally, at all times.
As we look at each one of these, these all are defined as a platform. Each platform, as every good platform does, there’s battles for dominance in that platform. From the desktop, Microsoft and Windows clearly won the desktop battle. And then in the mobile space, Apple has clearly won the mobile battle. And in the cloud, being in Seattle…I hear different opinions of who’s actually winning or has won the cloud battle, but Amazon right now is definitely in the lead on that one.
And one of the things that that shows us is that dominance on the platform does not translate into future success. Microsoft wasn’t able to take their dominance on the desktop and translate into their mobile platform. And we’re looking at Apple right now, who’s trying to use their dominance in the mobile space and translate that into Siri on all their devices, are they gonna be able to be successful in really making that leap and translating that dominance to all of their products.
Brian: History tends to say no.
Sean: History does tend to say no. So, what are kinds of obstacles that we have today that we are gonna overcome? I know in the previous presentation, there was a couple of jokes about the election next week, but one of the things I’ve thought about pretty heavily is that whoever gets elected next week, identity is going to be one of the things that they’re gonna have to address during their administration.
We see it in the news just about every day. EU has gone through a revamping of their privacy laws. There’s a lot of things about identity that, really right now, is not very managed. If you think about if you go shop on Amazon, you have an identity that you go. If you log into Facebook, you have a different identity. If you go into Google, you probably have a different identity. If you use LastPass, that’s something to manage all these identities that you have. And when we’re starting to blur the lines between the digital world and the physical world, we need to really think about it from a single identity perspective.
So when I walk into a hotel, I should be able to look at my phone or my watch or have something that tells me, “Well, your room is room 220,” and I walk right up to the room, get to the door, the door unlocks, I walk in, and if it’s enabled to have the ubiquitous OS, then it knows I’m in the room and I have immediate connectivity to my assistant to be able to figure out where I need to go next. And so there’s a lot of things with an identity that’ll have to change in order for us to really be able to have that digital proxy.
The other thing is that in the world now, data equals dollars. Data that we have as we browse, it’s clicked, it’s tracked on a level. I know that I’ve even put in some of those tracking systems myself and everything is tracked that we do. And the organizations hold onto that data because it’s viewed as a monetizable asset that they’ll be able to use at a later time.
And when organizations are using that data for their own financial gains, it doesn’t really build a lot of trust in how I feel they’re going to use my data. I certainly don’t think that they’re acting in necessarily my best interest because it has become a revenue stream.
With that also being a revenue stream, as you looked at the conversation that Brian and I had, it was a coordination of a lot of different services, and there was a lot of – we’ll call it – intimate information about what my favorite restaurant was, how would a machine know what my favorite restaurant is, other than having access to not just pieces of my data, but all of my data and being able to analyze it and know what my preferences are and where I wanna go.
So, this is what life with ubiquity looks like, and so all I have to say is, “Are there any questions? Because he looks pretty comfortable. But the real thing here is that with this, there’s gonna be winners and losers, very similar to…in the mobile space, you have the works with Android, works with iOS. As the ubiquitous OS really starts to catch on, everyone’s gonna want their apps to be able to work with the personal assistants because now we’re adding value to the people who are using the personal assistants, making their lives easier. And so I would ask a lot of organizations, “Is your icon gonna be on this picture?”
So the, you know, I kind of talked to Brian and I said, “Well, we’re kind of using a crystal ball here. How do we know if what we’re doing is actually…?,” if we’re on track and if we have ways of testing ourselves to see if this is what’s really happening. We’ve broken things down into four different phases that we’re seeing things occur in. One of them, the messaging APIs, we’re already there in a lot of cases. With the example we have here, I can go and book on American Airlines and I can tell it “Make sure an Uber is there to pick me up when my flight lands.” That’s really interaction between two independent services where they basically have an agreement with each other where they can send messages back and forth to each other.
The next one is the ad hoc query APIs. This is the really interesting one because this is where we are right now. Right now, it’s really more of a command interface. If you have an Echo, an Xbox, your iPhone, the new Google Home device, what these are is it’s really a command interface. I say, “Alexa, play this song,” “Google Home, turn on my lights,” “Siri, what is a restaurant near where I am now?” It’s very much a trigger word, and then I give it a command. That’s really not how we interact. I know that if I did that with Brian, he’d probably say, “You’ve gotta move off my team now.”
But Google Home is actually introducing something new in that where it’s becoming context-based, so I could have a conversation with the device. So a good example would be, “Google Home, I’m coming to New York. And I’m a huge baseball fan, and I know they’re not playing tonight, but are the Mets playing while I’m in New York?” The device would say, “Yes.” “Are there seats available for the game while I’m there?” The device would say, “Yes.” “Can you get me two tickets that are $40 a piece?” So that’s the type of conversation where there were three questions being asked that were all linked contextually and the device is able to keep up with it. Now, Google Home isn’t quite there yet, but they’re the first ones really taking that dive into that.
And then the next one is federated cognition. This is where we have a group of services working together. And what’s really different about this one is that now we’re starting to talk about platforms. We’re starting to talk about a mechanism that can manage the orchestration of all these different services that we may need to talk to, similar to what you saw on the slide that Brian had showing Facebook and OpenTable and Uber and all the different services we may need. And this is where, as we move to the ubiquitous OS, where the platform battles are gonna be fought because now we’re talking about services that are allowing an orchestration engine to allow all these services to really talk to each other.
And in this example, very similar to the first one, but it’s, “Siri, have an Uber pick me up after my American Airlines flight lands in Seattle,” where what’s actually happening here is that now the platform is orchestrating all those calls and then being able to give me an answer.
And then the last step will be the ubiquitous OS. And the best way I can really describe how that would be using a virtual personal assistant, which is really gonna be a key face of the ubiquitous OS, is in this example. We say, “Schedule a meeting with Brian in my favorite restaurant and have a car pick me up when it’s time to go.” It’s very non-specific. The ubiquitous OS knows that I like to use Uber. It knows that my favorite restaurant is Javier’s. And the virtual personal assistant then takes care of all the details.
Brian: All right. We’ve reviewed kind of the future through our eyes and what we see down the horizon. We’ve talked about the tech that’s gonna be involved in taking us there and now we wanna kind of drill down into the challenges that organizations are gonna face.
Sean indicated before we’ve helped many of our clients deal with the mobile revolution and we have many clients through taking mobile-first strategies. Well, any time you’re talking about a technology wave, that wave is a point in time. This is a point in time that we’re talking about where mobile-first won’t cut it in the future, where things that happened in the past that worked for marketing organizations wouldn’t work moving forward.
So just try and imagine this virtual personal assistant that’s making – I’ll say – my standard, run of the mill buying decisions. When I have a machine that is making my buying decisions for me, marketing to humans doesn’t quite work anymore. You’re now having to figure out how to market to machines in order to get your product to be the one that’s bought because I’m completely out of the loop in even making that buying decision. I have granted my trust to my virtual personal assistant to do that and so there are quite a few challenges when we start moving into that model.
The first one is about influencing algorithms. Now, this sounds kind of…well, it’s exactly what you’re gonna need to do in this model where we’re talking about influencing buying decisions, but it’s not completely foreign to marketing groups. Marketing organizations have been trying and attempting, mostly successfully, to influence algorithms for at least the last 15 years. From search engine optimization to pay-per-click aggregation, that’s there.
The difference is that when you’re doing search engine optimization, you’re basically trying to game the algorithm in order to get human eyeballs in front of your product. We’re now talking about the next step where you’re needing to game the algorithm in order to make a purchase happen, so just kind of that next phase. So, when we talk about investments, investments in continued algorithm optimization for marketing organizations will be one critical aspect.
Additionally, the sentiment is really high on the list. This is not new either. Marketing organizations across the country are very focused on making sure that they have good sentiment out on social networks throughout the internet, but the difference now is if you have machines making buying decisions, that sentiment becomes even more important because algorithms will be used to not kind of review subjective analysis as a human and make a subjective decision, but literally to score the sentiment that’s out on these social networks and make a scored buying decision. So sentiment becomes even more important in that model. Even when you’re doing sentiment against human ratings, those human ratings are gonna be scored by a machine to make a buying decision by a machine.
But let’s go one step further. With the Internet of Things, now we have another contributing factor to sentiment. Now we can have devices that are reporting on uptime for products that are talking about repair schedules, breakdown and there’s no subjectivity to them. These devices are reporting on how well the products are doing and the machines are reading that device data and making a decision on whether or not that’s a product that they’re gonna continue to buy for their human counterpart. So you can see that it’s kinda just moving more in that direction in all phases.
Additionally, Sean talked about this kinda platform battle, and this is really critical. It was critical in the beginning stages of the mobile wars, and you can’t read the crystal ball and know who the winner’s gonna be. You can make some informed guesses, but really what’s gonna be required when you as an organization are making some big bets on the ubiquitous OS, you’re gonna need to have good strategies to kind of pick the players at the time that are looking most dominant to do some proof of concepts there to adjust and pivot when necessary as the market changes. This will be a fast-moving category that won’t have winners for quite a long time.
Additionally, from a privacy perspective, these virtual personal assistants are gonna serve as the new ad blockers in this world as well. If you think about it, we’ve all experienced the event where you go shopping for tennis shoes online, and from that point on for the next month or two, all you see in your Facebook feed is tennis shoes. They know you were looking for tennis shoes.
But what helps with a virtual personal assistant that really knows your whole entire space, is she or he also knows that you already went and bought shoes at the physical mall and she’s gonna prevent those shoe ads from popping up. So, a marketing organization, you have to be aware that just because you think you’re pushing ads to your humans, you may not be, because that virtual personal assistant, just like the real life human administrative assistant prevents the CEO from getting meetings with vendors that aren’t valuable to him or her, the same goes for these virtual personal assistants. They’re gonna be serving as that gatekeeper and making sure that you only get the relevant marketing material and ads that come along with that process.
So what are the implications of all this? To wrap up, what we’re talking about is algorithms continue to be very important in terms of understanding and influencing human decisions and machine-based decisions. Sentiment and consumer reviews will also…you will need to continue to invest heavily in making sure that you have the right ratings and evaluations out there for your product, and there’s a lot of espionage that goes on in sentiment ratings and you’re gonna continue to have to deal with that and make sure that the algorithms that are reviewing those sentiments look upon your product favorably. And then really kind of hedging your bets on the right platforms, making sure that you have teams that are kind of monitoring where these platforms are headed, which ones are winners, which ones are losers. It’s gonna change for probably every six months, just like it did in the mobile space, as you’re monitoring and keeping track of the industry.
Then from a tech perspective and in terms of investments, APIs and identity are really the key components there. A great way, if you’re in a marketing organization, to figure out whether or not you guys are on track and we actually do this with our vendors when we’re even evaluating SaaS applications and these things is to ask, “Is there anything I can do in the application you’ve given me that I can’t access via machine via an API?” Usually, they look at you like, “Oh, they caught me,” and the answer is no.
You look at the smarter organizations like Amazon, Amazon has a mandate that you get fired if you release anything as a product that can’t be accessed as an API because for the layman, an API is a way for a machine to do what you might do through an interface. And if we’re moving to an environment where the machines are always triggering the purchasing decisions that are triggering reservations, then you’d better make sure that it doesn’t require a human sitting down to a web browser to do what you need to do. It needs to be able to be accessed by a machine.
So, with that, I think we’re really excited about where this is headed. This seems really far off, but I guarantee if you just start opening your eyes to where things are going, the technology is moving very quickly. And this is not like a decade away by any stretch, these are over the next few years that you’ll see this really take hold.
So, thanks.
Audience 1: I have a question.
Brian: Oh we have questions? Okay. Yes?
Audience 1: [Inaudible 00:30:21] talked quite a bit about customer context and understanding customers, but that’s one aspect of what you guys are talking about here. The other aspect is how do you understand the customers. So it’s ingestion and then that understanding Like, when we went to Microsoft, we saw [inaudible 00:30:43] them talking about this. What’s your take on what technology stack is necessary to kind of bring this stuff to fruition?
Brian: Yeah, I think it’s a good question to make sure that when we’re talking about investing in technology stacks that we’re level-setting. We’re not saying you need to go deploy your development teams on creating a natural speech recognition platform. The platforms, that dominance battle is gonna be won. Just like you wouldn’t be developing a mobile OS.
So when we talk about the investments that you wanna be making, there’s some things that there are…there’s certainly experimentation you can be doing right now that gets you to, if we went back to Sean’s slide here…you most certainly should ensure that you’re baseline API-accessible in everything that you’re doing, right? And from a command interface, that’ll come and go, because that really just…some of that natural language processing is gonna be handled by the platform. The path to those ones down below are really vendor alliances between different organizations. I can guarantee you that Uber probably cut a deal with American Airlines where it’s like “You’re gonna work with us but not Lyft.” That’s no different than other types of alliances that you start to see forming. I don’t think we’ll see that at the platform level where they’ll say you can be on Siri but you can’t be on Cortana or something, but you may see that early on. There’s a little bit of siloing that people try and commit people to until they realize that’s not a, long-term, good strategy.
Audience 1: I did have a similar question or a follow-up question just because you have this slide up. But one of the things that I think will be the most difficult part is kinda getting collaboration across competing platforms which will ultimately be necessary to make it really ubiquitous. Do you really see multiple ubiquitous OSs that you kind of opt-in, like in ecosystems? Because that’s kind of the world we live in today. You’re either Apple or Google or Amazon and unless they all kind of talk to each other and actually make a pact to collaborate, we’re always gonna be siloed off and kind of…
Brian: I would imagine, and Sean can elaborate on this after I say this, but Sean and I in that example had two different OSs that we’re going on. Allison knows my data, right? So Tom knows Sean’s data and we agree, “Hey, you guys can exchange information between each other for the benefit of both of us working on something?”
Sean: And we actually kind of talked about that from the perspective of, okay, we’ve got the battle is raging between Apple, Google, Microsoft, and Amazon.
Brian: A little bit of Facebook.
Sean: What?
Brian: A little bit of Facebook.
Sean: Maybe a little bit of Facebook. But think about, you know, we have Captchas whenever we go to log in to make sure that we’re human whenever we go to certain websites. So are there gonna be agreements, like, with Apple? Say, for example, Uber signs an agreement with Apple and I wanna use Uber, but my ubiquitous OS or my personal assistant is actually running on Amazon. Are they gonna be able to prevent me from being able to communicate?
And what we actually landed on was that’s actually probably… will create a really interesting paradigm, because then the virtual assistants will have to learn to be more human so they can act like they’re human and we actually then can go in a totally different direction from there. But I think that that type of thing is gonna happen where the platform’s, during the battle, there’s gonna be alliances made and there’s gonna be incompatibilities. But as it always does, when it ends, there will be one standard that kind of comes out of it even though there may be multiple providers.
Audience 2: [Inaudible 00:34:47] point. I thought it would kind of interesting in terms of my own life in digital corporate investments and where I’ve got kind of like you could just lock in [inaudible 00:34:58] Apple. But, like, when it comes to Pandora or Yelp or any number of applications, I pretty much can access them with Alexa, on iOS, [inaudible 00:35:07] or whatever. So now I find myself now making decisions on, like, where I started to try to, like, invest in more portable insights for me to be able to, you know, access them. It’ll be interesting to see how Google systems respond to that. [Inaudible 00:35:23] acquisitions and things like that to try to build that lobby.
Brian: I think what you may see as well is there’s gonna need to be some kind of wall. Pandora may not trust – and shouldn’t trust – that they can give me my data to where I can hand it over to Spotify. But if they can trust that Allison can use that data for the betterment of me without handing it over to a competitor, and there will probably be some legislation around, actually, ownership of data. They try that now, but who owns the data that you have when you have those services, and there’ll just be more of a demand for it when they can actually see the value of a machine actually harnessing that. But early days, it’s gonna be really kind of close to the chest in those services.
Audience 2: That’s a good point. Alexa will trigger a Pandora playlist [inaudible 00:36:21]
Brian: Right.
Audience 2: So just a real quick lesson then. You didn’t even mention kind of chat bots and message bots.
Brian: I actually had that on there and I just forgot to mention it, but yeah. So, yes, what we did talk about a little bit, and I think this is critical, is I would be really skeptical if I said the next wave of this is all voice-based conversations, because none of us always need to operate in that way. Depending upon the task that you’re doing, if you’re sitting down at a desktop, you’re gonna annoy your cubemates and everything else if you’re having to have conversations to compute a spreadsheet.
But taking advantage of the medium that you’re in, if I’m texting, which…you know, that’s the other challenge that it would have, right? Like, how often am I ever talking on the phone anymore, right? Like, so text becomes really important there, and we’re already seeing that with message bots. The real key thing in message bots being that, for Sean and I, we’re having the exact same conversation over text. The exact same thing could be happening, but all of the sudden we see a text message from Allison popping up instead of a voice. Other than that natural language processing, actually, the complex part of natural language processing is not determining the words that I say. We’ve got that dictation down. It’s really understanding the meaning behind them. So that really has to be figured out, whether or not it’s a message bot or in voice, and you’ll see Allison and Tom and others span both the messaging and the voice.
Audience 2: But in the message context, it’s almost a little more voyeuristic because it’s like interpreting kind of the context. It’s not even necessarily [inaudible 00:38:03], saying, “We’re having a conversation around dinner.” And all of a sudden, Yelp pops up and says, “Hey, [Inaudible 00:38:09]
Brian: As long as they don’t become the Clippy of the new modern world and say, “Looks like you’re thinking about dinner,” and you’re like, “Get away, Allison. We’re just having a conversation,” then you’re good.
Audience 2: What we saw also at the conference was that the technology, that the text to speech and speech to text are becoming so good that it can happen so fast that they’re kind of the same thing, right? So it doesn’t matter if you’re talking or if you’re typing, it understands both the same way. It can translate between the two of them extremely fast, right? Just like how we saw the translation between language where they translated it all with [inaudible 00:38:53] in less than a second, right? And that’s crazy, but it’s becoming the case that different communities of communication are becoming one and the same, kind of like what you were saying. It’s all integrated, so I think [Inaudible 00:39:11], right?
Sean: And, you know, again, looking at the Google Home, mine hasn’t arrived yet, but it’s the context. It’s really starting to see how context is gonna start playing a role in it because it’s great with these devices. My daughter who’s five years old can tell Alexa to play her, you know, Finding Dory or some Dr. Seuss book, but what I’m really looking for is I wanna be able to do something and I need the answer to this question leads me to this question, which can get me to my answer. So it’s really looking at how that’s gonna evolve at this point.
Brian: Can I make one more little quick… I’m gonna just say this. So, we did talk about dominance not being able to play into these new roles, and the one caveat I would say with that is Google, and the reason is because the amount that they already have under their belt would just really…having the index of the entire internet under them is really powerful.
We were walking here through Central Park and Sean wanted to know is FAO Schwartz still open over on whatever, and I get out Google, used voice to do it, to dictate the question of is FAO Schwartz still open in New York, and instead of just giving the list of web answers… If you’ve noticed, if you ask Google a question now, usually it will give you the answer. It doesn’t just give you a web page to find it. It said, “FAO Schwartz closed last July 2015.”
That’s not because it was prepared for that. It actually read the web page, understood what the web page was saying and gave me an answer to my question instead of a link to a web page. So that’s really… And Google has that over all the competitors right now and that really will give them a leg up in this new battle.