Amazon's Plan To Get The Power Of Alexa In Over 1 Billion Devices And The Future Of Voice AI

It's been just over six months since Amazon launched its Echo smart speakers and introduced its Voice AI Alexa to India.

To say Alexa is one of the most interesting things to grace the world of computing would be a gross understatement. If anything, Alexa represents a fresh new wave of voice-driven computing experiences that are making human-machine interactions more simple, enriching and intimate.

ALSO READ: Far From Perfect, Here's Why Amazon Echo Plus & Alexa Was My Most Memorable Tech Of 2017

However, you don't have to be a genius to realize that Amazon can't just bottle up Alexa inside Echo speakers. For Alexa Voice AI to flourish, for her to reach the masses, she needs to be available on every other product -- more importantly, devices not manufactured by Amazon.

How will Amazon increase the reach of Alexa? We spoke to Adam Berns, Director of Business Development, Alexa Voice Service - Amazon, to understand just that. Edited excerpts from our interview below.

Alexa Voice Service is...

"There's really three parts to Alexa overall," Adam starts to explain. "The first part is our first-party products like the Amazon Echo, Echo Plus, Echo Dot. But the second and third part is where third-party scale comes in."

Adam then went on to explain how Alexa Voice Service or AVS is what comes into play where Alexa gets leveraged into non-Amazon devices. Where a Sonos or Bose wireless speaker wants to implement and integrate Alexa into their product to give customers a whole new way of interacting with the device, for instance -- which is the second part of the scale strategy. The third and final part consists of Alexa Skills Kit, which lets you extend the platform -- much like an app store currently does -- and invite developers to take the Alexa experience and apply it to different scenarios.

"So literally," Adam explains AVS further, "it's a set of APIs where you could take a product like a wireless speaker or a computer or a car and then basically add the Alexa functionality to that product. And we have a royalty free license, where basically anybody can sign up and then build a product or an app that would have AVS functionality in it."

The range of devices for AVS

In 2016, there were 10 AVS devices available to customers. Today, Amazon claims to be approaching 100 Alexa-enabled devices from companies like Sonos, Anker, HP and Garmin. In all, over 12,000 devices that can be controlled with Alexa, from more than 2,000 unique brands.

"The type of AVS devices are pretty broad," Adam said in response to my question on different types of Alexa-enabled devices, before continuing. "So you can think of set-top-boxes we have deals like with Dish. In the automotive sector, we've done GM, Ford, VW, BMW, Jaguar, Fiat have announced they would be adding AVS into the head unit of the car itself. Then in computing, we've announced HP, in mobile there's Huawei, Motorola, HTC. On the wireless speakers front we have Sonos, Bose, Polk, Denon, Jabra, JBL."

It's a pretty broad range, Adam mentions, and because it's so new the AVS team is constantly learning. "I'll take the example of smart home," says Adam. "We're trying to build a smart home, we knew a bunch about how we wanted to implement a smart home solution powered by Alexa. But we also knew there was a bunch of stuff we didn't know. So we partnered with one or two lighthouse companies, like Philips for the Hue lights -- which are also sold in India. And together, we developed and defined aspects of the smart home domain that gave them early access into our technology, but it also helped them drive our roadmap. And on the way, we learned a ton. We weren't lighting specialists, but we were able to partner with them and learn. It's a similar journey with wireless speaker manufacturers and other device OEMs."

Why are device manufacturers integrating AVS

"I think the value proposition is really easy," Adam highlights, as our conversation continues, "and it divides into three parts."

The first part is simply giving a product that is hard to differentiate on its own a unique feature through Alexa voice integration. Adam explains this with the example of a router -- every router is kind of the same, has the same functionality, but now they add voice to it. Not only does that make routers suddenly very cool, it also makes them less complicated, less technical and more human. Imagine having the ability to just speak with your router and configure it easily? It's a no brainer, even I must concede that.

Adam then goes on to say, "The second value we deliver is that voice is just a must have. Like, if you're making a wireless speaker today, being able to voice control it, it's a must have. And the third is Alexa voice integration is a nice feature to have."

Adam explains the third point with the example of a PC, where its core function is productivity or browsing, but with voice integration it gives them another way of interfacing with the PC. "It adds functionality to a product that already has strengths on its own."

Controlling the Alexa experience and protecting user data

No conversation around Alexa or an Alexa-enabled device is complete without appropriate safeguards. How does Amazon ensure Alexa isn't misused on third-party devices? Adam explains, "If you're building a product that's going to interface with any of the music services, because the call will be through us, we want to make sure, obviously, that the rights holder are paid, and that that interaction is legal. And it's done in the right way. So we, if you have music playback, you have to go through our music certification. And then let's have agreements with each of the music services, Deezer, Spotify, and Amazon Prime Music.

"Then there's a set of requirements around customer privacy data notification. So for example, you have to be able to turn on the device physically, where the mics have to be pressed to turn them on or off, with a red light indicating that shows there is no power going to the mic. So there's things like that we write them into our agreement. And then in our agreement, we reserve the right to basically pull you off the service, pull your set of ideas off the service, if ever we detect that someone's violating the terms."

Alexa getting out of the home, becoming more public

As I converse with Adam, and start connecting the dots around Alexa Voice Service, I start noticing a problem. Because Alexa needs to be experienced to be believed, Amazon will never be able to achieve scale if people don't get to experience Alexa -- before they actually buy a device. Adam agrees.

Enter Alexa for Business. Adam explains, "Alexa initially was designed to be a consumer product that was tied to your Amazon account. So it was very personal. And the idea is to be in your kitchen or in your living room, it wouldn't move. And we realized that we needed to go way beyond that. Alexa for Business allows you to fleet enable anonymous devices."

So you can take Alexa devices and we put them in every hotel room, for instance, where you can control your window blinds, ask for the concierge and set a restaurant reservation do all of that from your Alexa-enabled device. That's one way of people getting to experience Alexa without really having to necessarily buy it, Adam mentions.

"Then we've had people who want to build kiosks for smart cities using Alexa. I think you'll see that coming up, we'll do more and more on the Alexa for Business side. And you'll see more sort of not only an enterprise focus, but what I would say is a generic device focus. Like it's not just your Echo tied to you, but it's going to be an Echo that can be used by multiple people."

Alexa and the future of voice computing

As we approach the end of our conversation, I ask Adam about how Alexa, or voice-based computing in general, is evolving the way humans interact with machines. While there's no doubt the intuitiveness of voice interactions, because it's still early days the scope of such interactions is limited. It's frustrating at times, too, as you learn the ropes of talking to a device rather than touch or type on it.

Adam tells me Amazon's take on this. "Our vision for voice is that it becomes an actual human interaction." And he explains this by taking me back to the early days of the World Wide Web where you did true Boolean searches, where search engines needed a certain syntax and operators to get results.

"You had to know how to do a search to get any response, right? If you didn't, you were screwed. But today you can pretty much go into your browser and type whatever and it will auto correct, guess your query and give you the desired result. The idea of being able to say Alexa, tell thermostat set temperature to 20 degrees start now -- we don't believe that's the way. With voice, we think you're just going to say Alexa make it warmer. Or colder. Because that's how a human conversation would be."

As our interview ended, I couldn't help but think about the potential of Alexa-driven experiences or voice-based computing taking off in a country like India, which has the world's largest population of illiterate people. They won't have to worry about typing on a keyboard, because they unfortunately can't -- instead natural voice-based digital interactions would help them cross the digital divide and open a world of possibilities.