With Amazon having such deep discounts on Echo devices over the holidays, and Google too, it wouldn’t surprise me if many of you have some new home assistants sitting around your house. As I mentioned in my last video, I just added Apple’s HomePod to my mix, so I now have always on speakers from Amazon, Google, and Apple in my home. And that got me thinking about the differences between them, the advancements they’ve made over the past year, and what’s to come in the future.
A Brief History
Siri was originally launched as an app in February 2010, bought by Apple, and then released as an integrated feature of the iPhone 4S in October 2011. In the beginning it received mixed reviews, but was praised for the potential yet to come. A potential that didn’t advance as quickly as many had hoped. She was slow to respond, wouldn’t give you the wrong information at times, and if she couldn’t do something, she’d crack a bad joke. Apple slowly worked on some of those shortcomings, but many people just became frustrated by her not meeting some very high expectations.
Three years later Amazon launched Alexa in November 2014, with the home automation features in April 2015. Unlike Siri, Alexa felt quicker to respond, and felt like she got more answers right than she got wrong. Where Amazon lagged behind Apple, and still does to this day to a certain extent, is her ability to be flexible and fluid with requests. You have to form you commands in a relatively specific structure.
And finally, the young kid on the block. Google introduced Google Assistant in May 2016, but quickly set the standard for how natural language should work with a voice assistant. Google became the gold standard and made Siri look like she was standing still, but didn’t have the depth of Amazon’s integrations.
So that brings us to today. The general consensus has been that Google is first, Amazon is second, and Apple squarely in third, but it’s a little more complicated than that. It really depends on how you want to use them. MKBHD did a great rundown a couple of years ago on voice assistants, so I’m going to riff of of his test questions.
We have some very basic questions, which all three should be able to handle:
- What’s the weather?
- What’s 15+5-3
- How far is it to London?
Then there’s conversational questions, which build on each other naturally:
- Who is the 45th President of the United States?
- How tall is he?
- Where is he from?
- What is the capital city of Russia?
- What time is it there?
And of course with smart home speakers there’s music:
- Play Daft Punk Get Lucky
- What’s the song that goes, “Like the Legend of the Phoenix.”
- Turn off the office lights
- Set the home temperature to 70
All three handle the basics well, but there’s a big difference when it comes to integrated services. When it comes to music, Amazon offers the ability to use Amazon Music, Apple Music, Spotify, Deezer, Tidal, VEVO, Pandora, Sirius XM, iHeartRadio, TuneIn, and Gimme. Google supports Google Play Music, YouTube Music, Spotify, Pandora, and Deezer. And Apple … well … it supports Apple Music directly on the HomePod. That’s it. Anything else will have to stream over Airplay from your iPhone, iPad, or Mac, but that’s not as bad as it sounds. I covered this in my review of the HomePod, but with Siri Shortcuts, any developer can add Siri control to their apps. I can use Siri directly on my HomePod to start playing my favorite podcasts from the Overcast app on my iPhone, no matter where my phone is in the house. If services like Spotify start adding that functionality into their iOS apps, then you could run Spotify the same way. But for now, we’re still waiting on those services to make those updates.
There’s too many home automation integrations to list out, but the major platforms like Philips Hue, Logitech Harmony, Lutron Caseta, Smartthings, and more work on both Google and Amazon. Google has been catching up quickly to Amazon this past year. Apple is the one that’s lagging behind with it’s HomeKit support, but that’s started to change since Apple loosened up their HomeKit requirements in 2017 to get more developers on board.1 Previously Apple required a special authentication chip included in all hardware, but now it’s all handled with software. The change has allowed some manufacturers to retroactively add HomeKit to existing devices, which has helped to spur the growth. And if you’re comfortable running a little code, you can install an open sourced project called HomeBridge on any computer that will allow you to integrate pretty much any device into HomeKit, like a Logitech Harmony remote control.2
While voice assistants have made big improvements over the past couple of years, we’re still in the early days of voice interaction development. Conversational computing will never replace our current forms of computer use, but they can augment it and add new interactions that weren’t possible before. It’s hard to watch Star Trek or Tony Stark talking to Jarvis in the Marvel movies and not wish you had a personal assistant like that. What we have today is a long ways off from that, but it can still be very useful.
The YouTube channel, TechAltar has a good video on “Is voice the future of computing?,” which I recommend giving a watch (I’ll include a link in the description). In that video he makes a great distinction between the strengths and weaknesses of voice, phone, and traditional computers. He points out that voice is great for quick convenience, but lacks context and is essentially a one way form of communication. You ask, it responds. You can’t have simultaneous communication like you can on a device with a screen. You have to take the information it gives you in the order in which it’s given, you can’t scan, jump around, and prioritize what you want first. Again, interesting video and worth a watch, but I’d add another big point … discoverability. Users usually have affordances with real world objects or software that hints as to how that thing is meant to be used. A simple example would be a webpage where you see the content is cut off at the bottom. That hints that there’s more to be seen below and you’ll instinctively swipe or scroll the window see what else there is. On these audio only voice assistants, there are no good affordances on how it’s meant to be used.
So when someone tries to ask a voice assistant to blindly do something in the hopes it can actually do it, and that request fails, it’s incredibly frustrating. The result is that most people will probably just give up and stick to what they know works for sure. As companies figure out these usability issues, come up with ways to hint at features and gently train us on how to use them, they’ll become more and more essential in our daily lives. Reminding us of appointments as we’re getting ready for work, or as Google demonstrated in Google IO from 2018, the ability to actual schedule and make appointments for you over the phone.
One of the biggest challenges to voice assistants is going to be dealing with the “creepy factor.” Many people got creeped out by Google’s demonstration, which on one side is an amazing technical feat, but on the other seems wrong to have an A.I. fool people into thinking they’re talking to an a real human customer. This is a massive gray area that needs to be navigated carefully for user acceptance, but the technology is clearly evolving quickly and will be very useful.
Which voice assistant is right for you? As you saw from the demonstration, they can all do pretty much the same things. Siri and Google are more natural to talk to, but they all have the same core skills and abilities. Alexa is no slouch and has the largest set of skills and integrations of the three. Based on where things stand today, my recommendation would be Alexa for someone who’s deeply into home automation, as well as wanting the most flexibility for music streaming services. But if you’re all in on Google and their music streaming services, a Google Home is obviously the perfect fit, and still gets you a solid list of home automation integrations. And if you’re pretty much an all Apple household and use Apple Music, then the HomePod is a great choice and one of the best sounding smart speakers you can buy.
And one final piece that everyone needs to consider: trust. Which of these companies do you trust with your personal data and security. These are always on listening devices, which record and transmit your voice requests somewhere when their trigger words have been called. There’s a quid pro quo happening where you agree to give up a little privacy, so companies like Google and Amazon can create advertising profiles based on data it gleans from your use of the product, and in return you get to use products that offer you value … for “free.” There’s nothing wrong with this business model, but you have to do your research and go in with your eyes open. Of the three Amazon has had a few recent lapses and data breaches3 the cast some doubt on how they’re handling user data, the same goes for Google.4 Apple is the one that is best positioned to protect your data at the moment, and they use this as a major selling point for all of their products.5 Siri anonymizes and encrypts everything you ask it. Apple can’t associate any data it collects back to a specific account, which makes that type of data collection completely useless to an ad network driven company like Amazon or Google. If you want voice control features with as much privacy as you can get, Apple is your best option at the moment.
But for straight up voice features, you really can’t go wrong when it comes to the core functionality on any of these, but just make sure you’re selecting a platform that supports all of the major services that you like to use, and that you trust with your personal data.