My experiences with four digital assistants

digital assistants image04

By Kevin Casper

Have you used a digital, voice-activated assistant? I’ve used four. I wanted to share some of my experiences, insights, and comparisons of them.

A quick look at who they are –
Alexa – Developed by Amazon exists on their Amazon Echo ecosystem
Cortana – Developed by Microsoft, exists in their Windows 10, mobile, and soon, Xbox One environments
Google – Unsurprisingly developed by Google, available on Android devices
Siri – Developed by Apple, available on iOS and soon, macOS (previously named OSX)

Before going into more detail about each of them specifically, I want to cover that they all can handle things like math calculations, add reminders to your calendar if you’re using the right calendar, and give you Wikipedia answers or refer web search links for you. Now that that’s out of the way…

digital assistants image03

Alexa

In my opinion, Amazon’s Alexa is currently the most impressive, but most restricted, of the assistants I have used. Alexa only exists on the Amazon Echo hardware – wireless speaker systems that use Wi-Fi and Bluetooth connections to communicate with your other devices and the Alexa capabilities. The Echo hardware is rather good. The microphone can pick you up and understand you from another room if you project enough, the speakers do a solid job at high and low volumes, and its usability is fairly simple. They’re not cell phones, though, and require pre-setup wireless networks with internet connections in order to be functional, so you can’t just take an Echo device with you somewhere and continue to get help from Alexa.

Amazon offers developer resources for Alexa – the Alexa Skills Kit – that allows developers to tie their apps or hardware to Alexa’s controls interface on a relatively open basis. I have Alexa connected with my Spotify account. I sometimes commute to and from work by bike. After a long day of work, then wrestling with car traffic on a bicycle, I’ll arrive home exhausted. While changing out of my biking gear, I can simply say, “Alexa, play some jazz on Spotify” and she’ll start playing one of Spotify’s jazz genre playlists I can chill out with. It has been easier than digging my phone out of my bike bag when all I really want to do is collapse to the floor for a while.

Alexa’s ability to understand your request has, in my experience, blown everyone else out of the water. You simply call her name, or one of the other activation commands available in the Amazon Echo/Alexa settings, and speak your request. In my apartment, I have some Wi-Fi-enabled LED light bulbs by LIFX. By giving permissions between Amazon and LIFX, I can have Alexa be my light switch, brightness controller, and even change lighting colors through some voice commands. As a light switch, I can say things like “Alexa, lights on!”, “Alexa, turn on the lights please.”, or “Alexa, turn lights on” and in all instances, she’ll turn on my lights. It doesn’t sound like much, but when it comes to a machine recognizing commands, that flexibility of language can make all the difference in the world. Sadly, you always have to use one of the very few activation commands first, such as “Alexa” or “Echo” before going into any commands. It’s probably for the best.

digital assistants image01

Cortana

Halo references aside, Cortana is a solid partner in the digital assistant space. You can give Cortana a try if you have a Windows Phone or if you’ve given in to the free Windows 10 upgrade like I have. Cortana mirrors much of what Google assistant and Apple’s Siri have to offer – providing vocally-controlled access to its shared ecosystem. Cortana works best for these things when tied with other Microsoft services, such as OneDrive and Outlook Calendars, and Microsoft does a solid job of letting you know that. Otherwise, Cortana has solid control with your device. She can put your phone or laptop into Airplane mode if you ask, she can open apps, and she can control your media that you have playing, at least if Windows has the media control access to your Spotify client or media shortcut access to your browser.

Cortana arguably doesn’t quite have as many useful app connections as Alexa, which mirrors the Windows mobile app store’s position with its competitors. It’s not for a lack of trying, as Microsoft does provide developer resources for Cortana. It is worth noting that Cortana does have an impressive vocabulary understanding similar to what Alexa has. Microsoft has put many resources into machines understanding language, which shows through Cortana. Cortana doesn’t struggle in understanding what you’re saying, just has some hesitations in getting you what you’re asking for.

digital assistants image02

Google

Also known as Ok Google to some and Google Now to others, albeit the latter isn’t exactly correct, Google’s assistant stands out from the others by not having a friendly voice very actively behind it. Instead, Google’s assistant services are a combination of things, including the previous mentioned Google Now, that provides a range of assistive functions. Much like with Cortana, Google’s assistant is best utilized with a Google account on all of its services, such as Google Calendar and Gmail.

On voice control, Google provides a voice action method using the activation phrase “Ok Google.” Whatever follows will, for the most part, write into a Google search for whatever it is you asked about. When using the Ok Google service on a mobile device, you can also have it set some settings or open apps as well. What’s interesting with Google, though, as that it doesn’t speak as much as the others do. Most of the time, it’s a beep or blip response, or otherwise brings up the visual results of whatever it was you were searching for. It’s a bit less human in that regard.

Up until this point, Google has provided more interaction with its services in lieu of a voice controlled assistant. For example, when I get a flight confirmation from an airline, Google will nab the info from that email and create a calendar reminder about the flight, as well as even going so far as a mobile boarding pass on my Google Now feed on the day of the trip if it sees a flight check-in email confirmation. That lack of “human” interaction might change, though, as Google announced their plans for Google Home, which seems to be a competitor for Amazon’s Echo devices. Also, for consistency with the other sections, I should mention Google offers developer resources for its Voice Actions capabilities too.

digital assistants image00

Siri

Apple’s Siri is probably the public’s favorite digital assistant. It was definitely the first one to really take hold and flesh out the concept of having one for many users. Siri has provided a heavily documented, experimented, and criticized baseline for the convenience of having a digital assistant. From the snarky remarks about dividing by zero to figuring out the interesting ways of responding to more personal requests, Siri has shown up and grown up quite a bit in this space.

Initially, Siri provided access to most of the basic iOS apps, including calendars, alarms, phone settings, and app access. Siri has been a great fact finder and search method, but since then, there has been a ton of competition. With the expansion of “Internet at Home” devices such as my previously mentioned LIFX light bulbs, there has been more and more demand in getting devices to work together. Apple doesn’t have the best history of an open development environment, but in recent years, they’ve made some great leaps in that method. Just this week, Apple confirmed that it will be providing developer tools for Siri so that app developers will be able to make more direct use of her language understanding functions to run their apps. Apple will also be using much of what Siri has been learning to improve other components, such as its keyboard and autocorrect tools.

Conclusion

While these digital assistants aren’t perfect, there is still plenty of work being done on them. One of the biggest opportunities for all of them is back in that point about understanding language and vocabulary. Currently, none of these assistants have strong grasps of context. Alexa offers a couple of contextual commands, such as asking to “Tell me about <something>” followed up by a “Tell me more” after she’s finished speaking. Google showed off some deeper contextual language with its Google Home promotional video. Competition has done great things to the development of these assistants and it doesn’t seem to be slowing down. Once we get replicators figured out, I’ll do a followup with who makes the best Earl Grey tea.

Facebook Comments

About author