3PO-LABS: ALEXA, ECHO AND VOICE INTERFACE
  • Blog
  • Bots
  • CharacterGenerator
  • Giants and Halflings
  • The Pirate's Map
  • Responder
  • Neverwinter City Guide
  • About
  • Contact

3PO-Labs: Alexa, Echo and Voice Interface

Breakdown of the Google Home news by an Alexa dev

10/4/2016

0 Comments

 
So, Tuesday was the big "Made by Google" event, where they promised that all would be revealed. Leading up to it, the hype train was going full steam ahead.​ We here at 3PO-Labs don't have opinions about most of the announcements (alright, so maybe we have an opinion about whether we need another VR platform), with the exception of Google Home. On that front, we could talk for days...


We COULD talk at you for days, but we'll try our best not to. Instead, we figured we'd split up our outlook on the few morsels of information we got into a few categories. So, here's how we're gonna organize the thoughts:
  • ​What we learned
  • What we didn't learn
  • Areas where it's gonna be important for Alexa to have some sort of response
  • Opportunities for Amazon to make its value apparent or re-acquire the spotlight
We'll be looking at this specifically through the lens of developers building voice-enabled applications (there will surely be plenty of hot takes about the consumer side of things, so go hit up TheVerge or Ars for those). It's also worth noting that Google pulled the rug out from under us (developers) for the second time in a row, talking up how they'll have great integrations while simultaneously revealing only the most minimal nuggets of information about how those interactions will be accomplished. They promised another event (boy who cried wolf, anyone?) along those lines for December, so a couple months from now we'll find out how correct our assumptions are.

What we learned

Standard keynote, full of promises, some of which seemed worded vaguely on purpose. We managed to extract a few salient points, though.
  1. Google Assistant is magic. Or at least that's what they'd like us to believe. It certainly looked neat in their demo (although it seems like every time someone wants to demo a more intelligent system, their topic is "dinner and a show". Do people in The Bay just go out to upscale restaurants and concerts every night?). Obviously the value of something like this depends on how well it works across a broad variety of topics. If they just hardcoded in a restaurant integration, that's neat, but it's not a game changer. That said, even an increase in usefulness of a couple percent over what Google Now could already do would be significant. We'll see.
  2. No enabling Skills/Apps. The way this was presented was a blatant pot-shot at Amazon. It's no secret that skill enablement and usage per Alexa user is low, and a large part of that has to do with how much more difficult it is to use third part skills compared to built-in functionality.
  3. Cross-Platform. Google Assistant will be on various platforms, including Android and Google Home. No surprise here, Amazon is well down this path with Alexa Voice Service, and there's incredible value in having an AI be ubiquitous.
  4. Tie-ins with Google products. This is another one that should've been pretty obvious, but that doesn't detract from its import. Leveraging their products like Youtube and Chromecast are an obvious way to differentiate themselves despite being late to the game.

What we have yet to learn

The things we didn't get answers to certainly outweigh the bits and pieces we got. Some of this will start to trickle out as media outlets get their hands on demo units, I'm sure. But for now, here are a few key things that left us wondering:
  1. How we actually develop (if at all). With all their promises of a great SDK, and awesome integrations, they didn't actually give any information about how developers work with Google Home. In fact, they pushed that announcement out until sometime in December, a month after the product gets in the hands of users. Presumably they have AAA partners they are working with, but the Average Joe is out of luck until at least that time. It's worth noting that Mario Queiroz (The Google VP behind Home) has been fairly dismissive of Alexa's considerable accomplishments and of the value of its third party integrations, so it wouldn't come as complete shock to see them limit who could build for their platform.
  2. Will Home have its own content? Along those lines, it's important to note that the tight coupling between a GUI-first AI (Google Assistant) and Google Home might make voice-centric development for the platform unwieldy, if not impossible. There are a lot of Alexa skills (especially the "fun" ones) that only make sense in the context of that sort of voice-enabled user experience. Would those sorts of things even exist in Home's paradigm?
  3. How do users discover integrations. If indeed there is going to be no concept of enablement, and presumably everything will be invoked at the top level, how do users know what integrations they can use? And if they are still namespacing their integrations, how are they handling collisions without enablement? Should I immediately start squatting on words related to popular topics like "weather"? They came off as fairly cocky about this assertion, but it's a non-trivial problem to solve, so we're very interested to see what this looks like.
  4. Specifics. For all of their promises, they didn't really give many details other than little bits of fluff (like talking about what materials were used in the physical device).
  5. Push Notifications. One thing they didn't talk about was whether Google Home would support push notifications (which is by a long shot the most frequently requested Echo feature we see from brand new developers). Obviously Android already supports push, but it'll be interesting to see how Google approaches this problem, with rumors continuing to swirl that Amazon is finally going to relent in some form or another. Push notifications have serious privacy implications, and Google is already walking a fine line in the public's eye, due to the amount of data they capture about everyone. Google Assistant is going to toe the line even more closely, and an always-listening Google device that can start talking out of the blue might be too much for some consumers.

Alexa must be a burger, because it's catch-up time.

So, obviously Google's got some answering to do. But what does this all mean for Amazon, who has essentially a 2-year head-start? Well, to start they're gonna need to respond quickly to a few of the things Google is planning.
  1. Google Assistant. A quality search function is gonna be hard to match, and tying special operations into search is a neat idea. Amazon may not be able to provide feature parity in this regard, but it'll be important that they make strides.
  2. Top level invocations. If Google has truly solved this in an elegant way, then I can't see how Amazon can put this off any longer. They've done it (really well, I might add) once already with the Smart Home API, but we need more hooks to integrate all the way at the top. Requiring skill names for invocations creates a terrible user experience (or, more likely, no user experience at all, because users can't remember the skill name, so they don't use it). This is one of those cases where the developers and the consumers are pretty perfectly aligned.
  3. Music, Translate, etc? Google has some nifty first-party integrations, like Youtube (although, I expect this will be severely limited), and translate. It'll be interesting to see what Amazon pulls out of its bag of tricks as a counter. (It's not nation-wide yet, but Prime Now Restaurants could be a nice vertical integration).
  4. Cross-skill user info. You can bet your bottoms that the user info available to the Google integrations isn't going to be locked down the way Alexa user info is. There's no way all of the fancy "AI" suggestions work when all you get is a meaningless userid with nothing tied to it, that isn't even the same across skills. If Amazon wants "smart" skills, it'll be time to revisit that one.

Amazon's turn to pounce

Critiques aside, the Alexa team has a lot of low hanging fruit that they can pick. Hopefully they use this opportunity to make some significant moves and recapture momentum.
  1. Features. The number one thing Amazon can do to fight back is to keep delivering features. A few potential features have been called out above, but those are just the tip of the iceberg for all the possible crazy things Amazon could do. They may not be able to go blow-for-blow on search or translate, but their two year headstart and the massive development operation they've put together means that they can easily win enough of the other battles to offset any deficiencies.
  2. Roadmap. Google's cards are on the table. Now is the time for Bezos to take a peek at his own playbook, and switch from "infuriatingly secretive" to "a little bit more open", like he recently did with Blue Origin. I cannot state in strong enough terms the degree to which the lack of anything resembling a roadmap is hindering (third party) development for their platform.
  3. Dedicated developers. Amazon's hidden weapon is that it has an army of third party developers who want to build cool things. Google's demo  didn't do enough to cause a mass exodus of the platform, but even so if Amazon doesn't step up it will continue to lose some of its most ardent Alexa developers and community supporters. We at 3PO-Labs don't speak for the entire range of Alexa devs, but there is certainly a sizable portion of the dedicated development community who feel voiceless and disenfranchised (blog post on that coming soon). Seriously, throw us a bone.
  4. Loosen the rules. Along those lines, Amazon 100% needs to cut its developers some slack. Right now nobody is getting paid (in fact, we're almost all paying Amazon in AWS fees for the privilege to develop for them) on Alexa Skills Kit, and they continue to restrict the ways in which we can use the platform. Something's got to give.
All in all, Google's announcement was not the bomb that some expected it to be, at least in the area of voice-enabled home assistants. Amazon is well positioned to continue their leadership in this domain, but they'll need to act decisively (both in technology and policy). And soon.
0 Comments



Leave a Reply.

    Author

    We're 3PO-Labs.  We build things for fun and profit.  Right now we're super bullish on the rise of voice interfaces, and we hope to get you onboard.



    Archives

    May 2020
    March 2020
    November 2019
    October 2019
    May 2019
    October 2018
    August 2018
    February 2018
    November 2017
    September 2017
    July 2017
    June 2017
    May 2017
    April 2017
    February 2017
    January 2017
    December 2016
    October 2016
    September 2016
    August 2016
    June 2016
    May 2016
    April 2016
    March 2016
    February 2016
    January 2016
    December 2015

    RSS Feed

    Categories

    All
    ACCELERATOR
    ALEXA COMPANION APPS
    BOTS
    BUSINESS
    CERTIFICATION
    CHEATERS
    DEEPDIVE
    EASTER EGG
    ECHO
    FEATURE REQUESTS
    MONETIZATION
    RECAP
    RESPONDER
    TESTING
    TOOLS
    VUXcellence
    WALKTHROUGH

Proudly powered by Weebly
  • Blog
  • Bots
  • CharacterGenerator
  • Giants and Halflings
  • The Pirate's Map
  • Responder
  • Neverwinter City Guide
  • About
  • Contact