Hacker News Reader

The example you gave asking about the the heart rate is my current gripe with Alexa. Alexa feels like I'm using an app via voice control, but if I could ask it questions and reliably receive answers like you described that would take it up to the next level for me.

I have no doubt that if Google wants it can get into the various partnerships, app addons, etc that Amazon's lead has given them, but if Amazon can't catch up to Google Home in terms of being able to understand and answer me without my having to get frustrated and just pull out my phone and search it myself, they're already ahead and as a user interested in these types of devices you already have me reading and considering switching. You're not the first person to point out that Home has better conversational understanding of voice commands in my searching post-reading your comment.

> Google wants it can get into the various partnerships

What made you believe Google can do this easily? When comparing Amazon and Google in their tracking record working with partners, I think Amazon always did a better job.

1. Amazon has the majority of their sales through 3rd party sellers. I can find source if you demand it, will be lazy ATM. 2. FBA is a huge success. 3. AWS's partnership seems a level ahead of GCP. 4. Amazon had a better relationship with international market, like in China, India, etc. And played a much more impactful roles in their presented market.

On the contrary: I really cannot find much success story for Google's partnership. They once acquired MOTO, then ditched it. They routinely release seemingly fancy and existing techs, and quickly abandon them in their newly shifted priority...

> Amazon can't catch up to Google Home in terms of being able to understand and answer me without my having to get frustrated

If you consider Amazon started with seemingly 0 prowess in AI (they actually have good data mining and ML stuff for their A9 branch, but that is so 0 compared to Google), and claimed such a lead in such a short time in terms of market recognition, then their speed actually is not an issue at all. Remember that Google's core strength perfectly matches the smart home applications, and Amazon's previous work in this area is pretty much blank. Such extraordinary improvement already proves Amazon caught up pretty fast.

You still have the key issue though. Amazon extends their business through Alexa directly. On the contrary, how Google directly extends their business through Home? Note how long does Youtube wait for successful monetization. If the same happens to Home, then during this time period, Amazon already achieved what they wanted.

When put into context, and correlated with past, the Amazon vs Google race in smart home, IHMO, favors Amazon.

That makes sense, and my reasoning for Google is they have money and they have users. The way they behave now certainly doesn't scream 'we're going to blow your mind with partnerships,' but I feel like if it was something they sought aggressively it would come somewhat easily. Pure speculation, nothing more. And if they chose to do so, with a superior interaction experience, there's the edge.

What you say appeals to my sense of logic, and I have a habit of buying the product that's cooler to me but the general populous lets it die (dreamcast comes to mind lol). I just don't think anyone should discount the AI here, and Alexa as an AI from the user perspective is not great. It really is almost just a voice controlled app from a use standpoint - if there's something more complex going on there, the user is unaware of it. I've never felt like I was using a virtual assistant with Alexa, ever. If Google can deliver a sense of actual interaction, that's amazing, that's the future to me. Basically for me the winner is the one who makes the AI that I can talk to like the Star Trek computer, and step one there is it has to answer my damn questions.

I don't understand how the list of "partnerships" makes sense. None of the examples (third-party Amazon sellers, FBA, etc.) are what I would think of in terms of services I want from a home assistant, unless the goal is to order stuff from Amazon.

Neither do I understand the difficulty finding examples of Google working with other companies: Chromecast, Android, the Play Store, and Google APIs are all examples, and much more relevant to an AI assistant at that.

In an effort to prop up pixel sales, Google is not rolling out the assistant to all android phones. If you read the latest blog post about the assistant conning to the shield and other devices, they have very carefully used language that excludes phones completely. They said the assistant will roll out to android tv, android wear and I think chrome books too. But not a word about it being released to phones other than pixel.

Risky move, IMO. If it fails, it has the potential to take the entire android ecosystem away from Google.

This is really disappointing. I was hoping Assistant would just be part of Google Now, but it looks like they want you to buy a Pixel, use Allo, or their own devices.

I sure as hell don't want ANOTHER messaging app when Hangouts is already dying off. And I was not impressed with the Pixel's price.

If it's any consolation, I own a Pixel and Assistant doesn't feel materially different from how I remember Google Now on my old Galaxy S5.

It works offline. There might be some AI chip in Pixel.

Interesting. That's an impressive sounding implementation detail, but as a user of several months I literally did not notice.

Unlikely given that you can hack an Android phone to have Assistant.

remember when google used to go on and on about how android was an open system and open wins and they had the moral highground saving us from the evil of apple?

what a joke.

What does Google's proprietary software have to do with Android being open?

it shows that all their talk of being open for principled reasons was bullshit.

It has made AOSP basically useless. Right now, you cannot compile AOSP and get a decent app suite. You need to include play services or develop your own browser, music player, video player, app store, maps, email client, etc. All of which is standard for any Linux OS.

Firefox, VLC, fdroid, osmand, k9. All open source apps on Android, you do you need to reinvent the wheel?

>Risky move, IMO. If it fails, it has the potential to take the entire android ecosystem away from Google.

I couldn't help but laugh at this. Google is the Android ecosystem.

I think it's possible Google had more in mind than just boosting sales of the Pixel - it's presumably much less expensive on the backend to support Assistant for a smaller subset of users, and Pixel owners are probably early adopters who will tolerate hiccups the wider market won't. My personal theory is the backend networks are being trained and optimized on a subset of users (pixel owners) and Google is making sure the product is worth the expense before hosting it for a much larger audience.

I think you're on to something. I wonder if they figure Pixel owners will be more forgiving than the average Android owner. Or perhaps just the lower volume of queries lets them keep a closer eye on things and squash issues as they appear.

>Pixel owners are probably early adopters who will tolerate hiccups the wider market won't.

I think you're confusing the Nexus with Pixel.The Pixel is targeted to consumers that buy flagship phones and do not tolerate being beta testers. The Nexus was more of a compromised reference phone to showcase the best of Android.

Here's a twitch.tv user that is streaming two Google Homes talking to each other...endlessly.

https://www.twitch.tv/seebotschat

Part amusing, part interesting in how the Google system is trying to pick up on cues in the speech.

At one point earlier today they were alternating the lyrics to Queen's "Bohemian Rhapsody" for some reason.

This is two cleverbots talking to each other though Google home, not google home itself.

I guess that makes more sense than random babble. Thanks.

Aah... For a second I was very excited to see the conversation !

I noticed that lots of companies are building Alexa in... but besides cars, the use cases haven't been particularly compelling yet, and would probably work just as well as a Skill or IFTTT channel.

And while Alexa has a lead in terms of Skills, the integration is so shallow that there is no lock in and I expect most vendor-backed skills to come to GH when the platform is opened up.

The fact that 5m people already have an Echo sure is something, but plenty more people have Cortana through Windows 10.

Google Home is Opened up via API.ai (now owned by Google)

I think that acquisition caught up google quite a bit on the SDK side.