“We never anticipated Alexa to have such a profound impact on society”

Jinoy Jose P | Updated on November 25, 2019

Miriam Daniel, Vice-President, Alexa Devices, Amazon.

Alexa has transformed the world of consumer technology in interesting ways. The smart voice assistant from Amazon has now become one of the most essential accessories of middle-class living rooms across the globe, enabling consumers to perform an array of household tasks and more with voice commands. Amazon’s Vice-President of Alexa Devices, Miriam Daniel is one of the minds that conceptualised and developed the iconic smart voice system. Daniel was in Chennai recently and spoke to BusinessLine about the evolution and growth of Alexa and the Echo speakers and Amazon’s plans for India. Excerpts:

How exactly did Alexa happen?

A lot of us grew up on Star Trek. It’s been 54 years since the iconic sci-fi show first hit our Planet and, in many ways, it has inspired scientists and technologists. Alexa is inspired by Star Trek, where characters used to control things via voice command. It showed an environment where you can speak into thin air and have an assistant act at your command. And that was quite interesting. Interestingly, this voice assistant wasn’t another computer-generated sound. In Star Trek, the voice sounded very human, it had a woman’s voice (Majel Barrett, if you remember).

So that was the big vision for us.

At Amazon, we used to do this exercise of writing a (mock) press release on a big idea we would want to see realised. Once, we wrote about a voice service that could play music for you, set reminders and could bring information to your fingertips. That led to us working on what would later become Alexa and Echo. Music was one of the biggest use cases for us so we thought about a speaker-kind of the device for Alexa. But the ride wasn’t easy. We had to invent a lot of new features for Alexa, starting from far-field voice recognition.

Even that time (early 2010s), phones had some voice services embedded in them. Many of them had dedicated buttons users could press and speak, but the experience wasn’t encouraging. A lot of people thought it was odd to use it in public. It wouldn’t understand you either. Now our challenge was to solve something that is not near-field but far-field. American living rooms are much bigger so we had to build solutions based on such premises. We needed fresh algorithms and technology to make that work.

So, we had to invent new noise cancellation methods, wake-word voice detection and a whole lot of new technologies. Inventing the wake-word technology was also important because a lot of background noise occurs in our houses, from fans and ovens to TVs. So the wake-word technology was also important. Also, it was important to build a natural language-based voice system for Alexa, because how Alexa responded to requests really mattered to a lot of people.

We had to be cautious about what Alexa could do, so initially, we came up with 13 skills for Alexa. That was five years ago, and these include bringing music, reminders etc. Now we have 100,000 skills across the globe. India has over 30,000 skills. We launched Hindi services a couple of months ago but we launched it with over 750 Hindi skills. Still, a long way to go.

But why build hardware for that?

We could have just built an AI programme and put it on phones. But that wasn’t easy because phones were owned by other manufacturers. Also, taking a cue from Star Trek, our idea was that you could stand anywhere in the room and talk to the assistant and get things done. You don’t need to pick up and phone, press a button and then give your command. So our vision was friction-less voice interaction. The next idea was to have a communal interaction, not a personal one like in the case of an assistant on a cellphone. We wanted it to be available to everyone and anyone in the room could use it. And that’s why we decided to build a unique device that you see today.

When we launched Echo in India, there were only two 3P smart speakers with Alexa built-in. i.e. Eufy Genie and Harman Kardon. Today, we have over 70 devices available from a price range starting ₹500. These Alexa built-in devices range from smart speakers, hearables and wearables, video and smart TVs to computing and smart-home devices from Indian brands such as Boat, DishTV, MyBox, Hero Electronics, iBall and leading international brands such as LG, Bose, HP, Fitbit, Motorola and Jabra.

Were you not concerned that there was no user interface to start with? I know many people have seen the arrival of Alexa and Echo as something that challenges the very idea of a GUI.

If you look at the evolution of technology, you can see that. When I went to college, there were keyboards and then came the mouse and then came apps and touch inputs, and voice followed. In fact, one of the most interesting things about voice is that you don’t have to teach people to use voice like they were taught how to use a computer or keyboard or a mouse. Voice was natural. It’s what you learn intuitively. So we thought about a voice-first model. Any screen would only be complementing the main service. We now use the screen to augment voice. Now we have Echo products that feature screens that work with voice. Voice and visuals is a new level of paradigm and that took us to a new level of user interface evolution.

Amazon was not a phone-maker then (the Fire phone came later). So what gave you the confidence that this idea of a standalone smart speaker would eventually click with the consumers?

Amazon has an appetite for big risky ideas. We were confident of this idea. However, not all big bets pay off. If you fail, it does not mean you lose faith in taking risks. I was in the Fire phone team and a lot of us moved to work on Alexa. So we knew what it meant to be taking on big ideas and taking risks.

How’s been your performance in In India?

This was the biggest sale event for Amazon Devices ever and we grew nearly 80 per cent over last Deepawali season. Fire TV and Alexa are a nationwide phenomenon now. This season nearly 40 per cent of Fire TV Stick and Echo devices sold were outside the top metro cities. Alexa Smart Home devices are also doing well. Tens of thousands of customers used smart lights this festival season. Customers loved the wide selection of smart TVs, lights, plugs and remotes that work with Alexa. More customers now set up and control their smart homes using Alexa – six times growth in smart home customers using Alexa in the last 12 months.

Are you looking at other vernacular languages in India?

Customers can choose from over 750 Hindi Alexa skills tens of thousands of developers are building for Alexa. In fact, our approach was to recognise that in India you cannot have a one-language approach so we decided to go with the mixed language approach. That was a first and was not done for any other country. There is a skill called Cleo. It is a crowd-sourced way to teach Alexa different languages. You can use the skill and tell Alexa and let her learn different words and meanings in different languages.

We have onboarded a lot of such training data. Hindi seemed like the next logical step. We’ll keep pushing the envelope. What has really caught our attention in India is the fact that Alexa and Echo devices are used by a lot of schools to teach kids English, general knowledge and other subjects. This is really inspiring and we will try bringing services and skills to enhance this process. We never anticipated this device which started off as a fancy geeky Star Trek-inspired tool to have such a profound impact on the society. Today, we are trying to make it more useful to students, teachers, people with vision challenges and so on.

Published on November 25, 2019

Follow us on Telegram, Facebook, Twitter, Instagram, YouTube and Linkedin. You can also download our Android App or IOS App.

This article is closed for comments.
Please Email the Editor