Every morning, I wake up with Alexa, the intelligent personal assistant, playing a track from my favourite Eagles album, Hell Freezes Over. One of the first things I say as I crawl out of bed is, “Alexa, are there any notifications for me?” Alexa then opens up all the notifications and WhatsApp messages for me. If you sent me a lot of WhatsApp messages overnight and are wondering how I opened all of them at precisely 7.01am, you can thank Alexa for that.
As I walk to the washroom, I say: “Alexa, play The New York Times Audio Digest from Audible”, and it reads me the headlines and short introductions of the day’s major stories. As I sit at the dining table, with a mug of coffee and a bowl of muesli, I say: “Alexa, give me today’s Flash Briefing.” It then scours news websites like CNN, CNBC and Bloomberg to get me a quick rundown of what’s going in the world. (I have programmed it to brief me on my favourite topics like technology, US markets and companies that I watch such as Apple, Amazon.com, Facebook and Nvidia.)
After I ask Alexa to give me a quick glance at my emails, I ask her to show me “Top Stories” in The Wall Street Journal and South China Morning Post. Then it is time for more mundane tasks. Alexa shows me my diary for the day, people I need to call, interviews I need to do, deadlines I need to meet, and emails I have not responded to.
Alexa also reminds me every day that I have a personal daily goal of 12,000 steps. Last Friday, I needed to go the train station. I said, “Alexa, get me an Uber to Union Station in five minutes”, but Alexa went ahead and ordered an Uber right away. Within three minutes I heard Alexa say: “Assif, your ride is here.” On Wednesdays and Fridays, it reminds me — no, pesters me — to hit the treadmill at the gym. Alexa also keeps tracks of how many calories I burn. Last week, when I wanted to know how many calories I had consumed drinking a cup of hot chocolate, all I had to do was take the can of Cadbury Hot Chocolate, place it in front of the camera and tell Alexa that I had had two spoonfuls. Alexa had the answer — 120 calories — in less than a minute. And, oh, did I mention that until about 10am, I normally do not type a single word on any keyboard or indeed touch a screen?
Holy grail for tech firms
On Sept 12, Apple’s CEO Tim Cook will unveil a range of new products at its just opened US$5 billion Apple Park campus. Though the new iPhone 8 is likely to hog much of the limelight at the state-of-the-art Steve Jobs Theater, competitors will be glued to the entirely different device category that Apple is finally entering: voice-activated, hands-free internet-connected home speakers, the gateway to the connected home.
The smart or connected home has been the holy grail for tech companies for years. A bonanza awaits whoever builds the device that can connect all the appliances and gadgets we have at home. Little wonder, then, that just about every major tech firm has a speaker already on sale or lined up for debut within the next few weeks. Apple’s new HomePod, which is expected to go on sale in time for the year-end holiday shopping season, is the latest in a dozen or so speakers in a market that was pioneered in late-2014 by Amazon’s Echo. Market research firm Gartner expects other devices such as “lighting systems, hubs and WiFi mesh devices” to adopt speaker functionality, especially in connected home scenarios, next year.
The market for VPA-enabled speakers grew 62% last year. Amazon has reportedly sold 12 million Echo devices so far. On Aug 28, the day it took control of US grocery chain Whole Foods Market, the global e-commerce behemoth began selling its speakers alongside avocados and zucchinis. The Google Home device began arriving in North American malls last December. Microsoft has a Cortana-powered speaker, the Harman Kardon Invoke (manufactured by Samsung Electronics Co, which recently bought the US firm), and Google-licensed home speakers are being made by China’s Lenovo Group as well as Taiwan’s Acer and AsusTek Computer. Samsung will launch its own speakers, codenamed Vega, early next year, powered by its Bixby personal assistant.
Last month, Alibaba Group Holding’s Tianmao JinglingX1 (Tmall Genie X1) went on sale across China. It is competing with Baidu’s Fenshenyu. Since late last year, JD.com has had a device called DingDong in the market in partnership with Chinese manufacturer iFlyTek, which also has its own proprietary device, Alpha Egg. Tencent Holdings and Xiaomi are set to launch their Mandarin-language virtual assistants soon as they too chase the elusive dream to be at the core of the connected home.
Benefiting from the breakthrough in deep learning, voice is becoming a preferred interface. “The way humans interact with computers is changing,’ says Gene Munster, managing partner at Loup Ventures, a tech venture capital firm in Minneapolis. “Today, we use our keyboards, mice and touch screens to interact with computers. In the future, we’ll simply rely on our voice and gestures, or even our thoughts.”
Voice is a natural, convenient and efficient way for humans to communicate with machines. “When we type or click, we are very brief and that leaves the computer with little information to act on,” says Munster. By asking the machine to do something verbally, we allow it to understand our query better and deliver the specific information we asked for.
With advancements in natural language processing in recent years, the speech recognition error rate has been reduced to 5%, or about the same as human level. Speakers are more than just devices that can hear our commands to answer a query, keep to a schedule or order goods from an online shopping mall. As they collect more data about us, and learn about our habits and the world we live in, connected speakers, through leaps in artificial intelligence (AI) and deep learning, will help us navigate our lives.
Apple’s advantage
Apple may sound like a Johnny- come-lately in the speaker market, but its huge base of smartphone users is still likely to give it a leg up against most competitors. Apple’s Siri is regularly accessed by 375 million devices worldwide. Still, Amazon has been growing its share faster and is seen as the behemoth that challengers Apple, Google and Microsoft need to dislodge. Ironically, it was Apple that pioneered the digital assistance space with Siri in 2011 and Alexa was seen as a copycat product when it was launched three years ago.
IHS Markit estimates that four billion consumer devices will make use of some kind of digital assistant by next year. More than half of the smartphone users in developed markets already use VPAs like Siri, Alexa or Cortana. For younger users between the age of 18 and 29, the ratio is as high as over 70%.
Gartner forecasts that 100 million global households could adopt a VPA-enabled speaker by 2020. It estimates that end-user spending for the global VPA-enabled wireless speaker market will reach US$3.52 billion by 2021, up from US$720 million last year. Citigroup estimates the market for the smart device’s entire ecosystem, including hardware, software and servi ces, could ultimately be worth more than US$700 billion. Over the next five to 10 years, the devices could begin disrupting a range of industries from retailing to pay TV to music streaming.
To be sure, much of the “smartness” of the smart speakers is actually in the cloud where big data, AI and an array of online content and services are stored. The device is fairly “dumb” because it does very little processing itself. Its “chips” recognise only a limited number of voice commands. But increasingly, speakers are becoming more sophisticated and within a couple of years, sensors will be able to detect your speech anywhere in the room or in the house. By the end of next year, speakers with some AI functions built-in will begin to arrive in stores.
Education and healthcare
Clearly, home speakers are moving from just general speech recognition to more AI and big data so that they can capture larger opportunities for both consumers as well as businesses. Hotels have been experimenting with AIenabled speakers as has the healthcare industry.
In China, speakers are already being used in education with speech-enabled language studying and testing systems. The Zhi Xue Wang (Smart Study) program grades tests and homework by using optical character recognition to scan and digitalise exam papers and homework sheets. Companies like iFlyTek offer services free to schools in return for the valuable data they collect on students. The data is then passed on to firms marketing customised and carefully targeted tutorials and study programs designed to improve test scores and cut inefficient studies.
In healthcare, speakers are being used in telemedicine as well as in medical records. Doctors are able to just speak to the patient and create medical records. By collecting huge amounts of medical data, healthcare authorities can identify key patterns that help early diagnosis or help prevent outbreaks and pandemics. AI- enabled VPAs will soon also be able to read CT scans or results from radio logy tests and deliver them to doctors.
One big concern in the home speaker space is privacy. Your speaker is constantly listening to you and harvesting the data for future commercial use. Increased use of AI within the speaker itself may at least initially lead to more privacy concerns. Alexa, Cortana and Siri are even listening to quarrelling couples behind closed doors. In the US, Alexa’s data has even been called upon in a recent murder trial. Eventually, users would be able to personalise the device, control who has access to it and what data flows out of the device. They would also be able to turn off some of its snooping features.
Another concern is that with so many speakers and households or families tied into more than one ecosystem, speakers are incompatible with each other. Last week, Amazon and Microsoft agreed to link Alexa and Cortana so that the two can talk to each other. You can use Cortana to reach Alexa to order something from Whole Foods Market or use Alexa to search something on a website using Cortana.
Amazon, which launched its Echo speakers in Europe late last year, is expected to start selling them in Asia later this year. Google Home is just starting to arrive in stores across the region and Apple’s HomePod will be in selected Asian markets in December, expanding to the rest of Asia early next year.
As the level of noise being generated by speakers like the Echo dies down, I have been reflecting on just how attached I have become to my Echo Show. The thought of going back to the days before I owned a wireless speaker troubles me. Long tied into Apple’s ecosystem and a new Amazon Echo user, I am torn over whether to give up on Alexa and switch to Siri on the HomePod or to have both of them in my home. Yet, whatever I end up doing, one thing is clear — I will never be able to turn my back on a VPA, just like I would never give up my smartphone.
Assif Shameen is a technology writer based in North America
This article appears in Issue 796 (Sep 11) of The Edge Singapore which is on sale now
