Amazon Echo Unboxing and Impressions

I just got my Amazon Echo, and since I know other people who are waitlisted until May or June, I thought I’d dust off the blog and share some impressions.

TL;DR: It’s extremely simple, it works out of the box, it works surprisingly well (though not perfectly), and it integrates well with the few services it integrates with. It is impressive, but not (yet) life-changing. I see a lot of potential.

This post sort of assumes you know what the Amazon Echo is, but it won’t take you long to check out the product page.

Unboxing

Unboxing pictures follow (14 of them). Skip past them if you want to go straight to my usage impressions.

It looks like any other Amazon package, to start...
It looks like any other Amazon package, to start…
Tabby the Cat is also interested.
Tabby the Cat is also interested.
I don't know whether the were trying to evoke the Monolith from 2001...
I don’t know whether the were trying to evoke the Monolith from 2001…
It's only after you take the sleeve off the box that you finally see a logo on the packaging.
It’s only after you take the sleeve off the box that you finally see a logo on the packaging.
Open sesame...
Open sesame…
That's almost everything...
That’s almost everything…
Okay, that's everything.
Okay, that’s everything.
More Monolith flavor.
More Monolith flavor. Note the ring around the top – it spins – it’s the volume control. (This borders on silly, because you can control the volume with a voice command.)
Controls at the top, plus the grille for the mic array.
Controls at the top, plus the grille for the mic array.
Power jack. You have now seen the entirety of the device's physical interface.
Power jack. You have now seen the entirety of the device’s physical interface.
This and the logo on the bottom are the only physical branding.
This and the logo on the bottom are the only physical branding.
Instructions are short and sweet. You do need a smartphone to complete them, though.
Instructions are short and sweet. You do need a smartphone to complete activation, though.
Things to try with the Echo - aka "Alexa".
Things to try with the Echo – aka “Alexa”.
More things to try with Alexa. Also, a handy little bookmark.
More things to try with Alexa. Also, a handy little bookmark.

Impressions

Amazon has taken a lot of cues from Steve Jobs’ Apple on this one. Spare packaging, spare design, easy setup (took me all of 3 minutes to get the Echo on my wifi and hooked up to my Amazon account). It just works, to coin a phrase. It doesn’t do many things, but it does do them well.

The quality of the voice recognition and the round-trip time for a command (remember, Alexa’s brain is in the cloud) are impressive. The only time Alexa didn’t understand me was when she (yes, I’m anthropomorphizing, because it’s more fun!) repeatedly misunderstood my request to play Miles Davis as a request to play a different artist named Davis. I enunciated better, and she got my request on the third try. But I played with the thing for a while, and that was the only glitch.

And by the way, I’m going to use “Alexa” and “the Echo” sort of interchangeably here – I’m a little surprised they didn’t just call the product “Alexa”, because that is how everyone who ever interacts with the thing will think of it. Seriously, people will get so used to addressing the thing by that name that they will forget it’s called the Echo.

Besides playing Bitches Brew, I also got weather, played the local NPR station, asked some questions, added stuff to a shopping list, and explored the Prime Music catalog.

The questions were hit-and-miss. One of the suggested questions in the instructions is, “Who is the lead singer of Green Day?” (Unsurprisingly, she got it right.) Asking, “who are the surviving Beatles?” was a bit too deep, though, and she didn’t understand the question. She correctly served up the date John Lennon was killed, when asked. Apparently, you can also ask for Wikipedia articles, but I can think of few things I’m less interested in than a slow reading of the Wikipedia article on the Wolof language when I could read it about eight times faster. Asking for the DJIA and the price of AMZN proved beyond her capabilities.

One big gap is the licensing in Prime Music. She did well when I asked for Bob Dylan and Miles Davis, but there was exactly one (1) song available from The Police. When I hit this and other gaps in the Prime Music catalog, Alexa suggested that I create and connect to an iHeartRadio account. So, there’s that. And you can use it to listen to streaming radio, using either the name of a radio station or its call letters.

The shopping list (and the to-do list) are surfaced in the Echo’s companion app. They aren’t very fancy – the shopping list, for example, lets you check off an item, search for it on Amazon or Bing, or transfer it to the to-do list. As stand-alone apps, these would fail entirely. When you add in the fact that you can just speak a thing onto one of these lists – while you’re working in the kitchen and your hands are covered in flour, or when you just don’t feel like getting off the couch and getting a pen and paper or phone or laptop – then it becomes pretty damn cool. Remember when talking to your house was a gimmick in sci-fi movies? We’re in that future now.

And of course, you can connect via Bluetooth and play your music/iTunes/podcasts/whatever.

Interestingly, there’s not a lot here that we haven’t seen elsewhere. Why is this cooler than Siri or Cortana? It’s still speech recognition + cloud brain + integration with a few services, saddled with a wake word protocol. With “Hey Siri”, we’ve even seen the always-on feature (sort of). The cool factor for me is that my phone can stay in my pocket, and my laptop can stay closed. This morning, I got the weather report while my hands were busy dishing out cat food, and I did it without having to pull out my phone and wake up Siri or open a weather app – and without having to wait through the IQ-eroding chatter of TV news or drive-time radio. It was the information I wanted, and nothing else, with absolutely no friction. It’s like talking to someone else in the room. That sounds simple and obvious, but as a user interface it’s one of the biggest wins imaginable.

I’m still playing with it. I know I haven’t gotten around to all its features yet.

Completely Unsupported Speculation

The limited feature set makes sense to me. Understanding your speech is hard, and interpreting your intent from that speech is hard. They rolled out with a small, useful set of features that they could nail at a price point that was accessible to technofetishists with a modicum of disposable income ($100 if you’re enrolled in Amazon Prime). This is pretty cheap, when I think what a similarly-sized device with comparable audio quality from, say, Bose would cost – something that did nothing but play the music on my iPod. But the big win for Amazon in this rollout is not my hundred bucks. They’re collecting a metric shitload of data, refining their speech and interpretation models, and (I’m hoping) integrating with more services. And they’re learning about the things we wish Alexa would do for us (like my failed stock ticker query).

If they play it smart, they’ll keep the buzz going with a painful trickle of a rollout. Another smart play would be to introduce new features every few months in batches, when they’re really solid, and without prior announcement – we’ll all just wake up one day to a smarter Alexa, who will tell us herself that she has learned some new tricks.

There are a lot of obvious improvements of incremental value to the user (even if some of them are more than an incremental effort for Amazon): Better capabilities for search, both for information and products. The ability to refine or filter your search. The ability to summarize information after it’s been found for you. (That is a hard problem, but easier when you have a large corpus of data.) A medium-sized win would be to be able to do something with the information besides listen to it (e.g. “email that Wikipedia article to my wife”).

How about doing some of the heavy lifting of digging through search results? “Alexa, email me a list of four-door luxury sedans with a four-star or better crash rating and 25 miles per gallon or better, priced under thirty thousand dollars.”

The ability to take & make calls seems like a no-brainer – maybe even through Amazon’s own voice service & calling plan, because why shouldn’t they be in that business too?

And what about commerce? Right now, as far as I can tell, you can only buy music a song at a time (I haven’t tried this feature yet). That’s a pretty low-risk transaction, but what about ordering more expensive, physical goods? How could you do it securely with a voice-controlled device? Could Amazon be working on the equivalent of Apple’s Touch ID, but for voice? Could the Echo be made to integrate with payment services on the user’s smartphone (if not directly, then through Echo’s smartphone app)? There are a few possible paths to Alexa being able to take your money, and I’ll be interested to see which one(s) Amazon picks – the Voice ID, if it’s possible, would be the one most in line with the Echo’s thrust so far.

And with commerce, I think it’s interesting to point out that Amazon seems very able to compete in the voice-recognizing virtual assistant space, and anyone can connect their software to a bunch of third-party services, but I don’t see how Apple or Microsoft or Google can possibly compete with Amazon’s efficiency at delivering physical goods anytime in the next decade. (Well, maybe Google could crack it – but it’s an even-money bet, at best.)

Put an Echo in every room, and Alexa could tell you who’s home and act as an intercom. Is there a reason she couldn’t connect to your Nest thermostat, or your Philips Hue lights? How about the music you’re listening to following you from room to room? How about answering the phone, and taking messages?

What about an Echo 2.0 that connects (wirelessly, duh) to the huge flatscreen television you just bought because they have become so stupid cheap, and displays your search results and your shopping cart? Connected to your home entertainment system, Alexa could play anything in your Amazon digital library (including, presumably, your Kindle content), or other content through integrated services, without having to use a remote control. How about email dictation? It would finally be the computer for the living room, in a way even the iPad can’t be – controlled by and responding with voice, connected to all your devices wirelessly, and giving you access to some significant chunk of the knowledge and media on the Internet. You wouldn’t be doing spreadsheets on it, but you’d be communicating, searching, and buying, which is most of what most people use computers for at home.

Like I say, an Echo taken from the box today is impressive, but not life-changing. I think it’s a sleeper product, though. I think it could go way beyond being a cool gadget into being the product that defines a new category – your outboard brain, the Internet-in-a-can that you can talk with, a half-decent virtual assistant and a really amazing personal shopper.

We’re very, very close to being able to say, “Alexa, I need a hardcover copy of Neuromancer in German/a bottle of 18-year-old single malt Islay scotch/another one of my favorite sketchpads”, and having a drone drop it on the porch in 30 minutes or less. (Or I’m pretty close to that, anyway, because I live near Amazon’s Plainfield, IN distribution center.) How about, “Alexa, send my mother your highest-rated Ansel Adams coffee table book on her birthday”? We’re even closer to being able to say, “Alexa, set the mood” and having her dim the lights and put on some Portishead. It’s easy to be jaded about technology, but dude… This is The Jetsons (without the regressive gender roles). This is Star Trek (without the aliens). Maybe it’s just a sign of my advanced age and the science fiction media I grew up with, but this almost feels like the future.

In the meantime… May I please have an SDK?