Is anyone else concerned that tools like Alexa are making the whole automated home thing too easy? I see it as creating a bunch of automation "experts" out of people who don't truly understand the whole feature surface/attack surface coupling. There are multiple tutorials on Amazon's website to learn how to implement the whole path, end to end. There are "smart" outlets that are actually pretty dumb, at least when it comes to knowing whats plugged into it.
Whats to prevent someone from connecting this to their smart door locks? Then if that person has a big party, and someone with bad intentions shows up, they can simply say "Alexa, at 3:30am tomorrow, turn off the alarm system, and unlock the doors". Or "Alexa, turn off the dialysis machine".
Sure Amazon might make it more secure in the future, but probably only after enough of these unintended consequences stack up.
I also understand that the average HN reader is probably not rushing out to introduce a single point of failure into their home security system. My point is that Alexa is essentially saying "don't worry about how it happens, just dream big about what awesome stuff might happen".
> What's to prevent someone from connecting this to their smart door locks?
August (the premiere door lock people) are very careful who they partner with, and wouldn't partner with anyone making an integration like this (they won't partner with my current employer for this very security reason).
I would argue that's a result of the way Homekit and Siri interact, and not the fault of the August Lock. August put the Homekit chip in the device.
Apple should really require a code along with the unlock command. It's also why Amazon asked some skill makers to remove unlock functionality recently.
It's harder to conceal a spoken command than a keypad-entered code. Spoken words are unavoidably broadcast to nearby surroundings, but you can mask your use of a keypad easily.
Amazon I believe recently asked a number of custom skill makers to remove unlock functionality from their code for this exact reason.
Could someone walk by my window and say "Alexa, set the thermostat to 90 degrees?" Sure. That would assume the attacker knows I have a smart thermostat, also they have to know the name of the thermostat in my app.
Also, my heat would have to be on for it to really matter, and at the end of the day all it is really doing is increasing my utility bill.
This Exactly.. I am really afraid to connect advanced skillsets like Uber on my Echo - for that there is no authentication methods on Echo.. To add to the rant, the Amazon Alexa is one of the worst android application with constant unresponsiveness and I am not sure they can deliver on a method where the authentication can be through the Alexa App for the advanced skills
I think this exposes a bit of a flaw in Google's rollout of their 'Assistant'; they only released it to Pixel devices early on which has alienated Nexus owners (or at the very least this Nexus owner).
I have both a Google Home and an Amazon Echo; the Echo shows some real strengths from its early lead -- a lot more connected and a lot of baked in bespoke interactions. The Home has Google's existing work on knowledge graphs built in and is surprisingly good at knowledge based queries (today I asked both 'what is a normal heart rate'; Google answered with some 'According to Mayo Clinic...' answer and the Echo said 'I dont know how to respond').
The thing CES this past week showed me is that the battle lines are being drawn. Amazon has a lot more partnerships rolling out; the only one I saw for Google was with the Shield, which is mainly because it's running Android TV.
I don't know which will win out. If I had my druthers, both would be available on 3rd party hardware -- I want to know that they are listening when I explicitly want them to listen, and I don't trust their own proprietary hardware to do so.
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.
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?
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.
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.
>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.
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.
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.
That Google is losing out to Amazon on this is to me largely attributable to Google always struggling with marketing. I constantly hear about Alexa. I didn't even know till now that Google has a new personal assistant your thing that's not Google Now. If I as someone who works as a programmer and reads HN every day don't know, how many regular consumers do?
It seems to me that Google mainly is able to get attention for their outlandish projects like car and Glass that go nowhere. That's probably because those products get tons of PR for free.
The flaw in Google's situation has always been its lack of a meaningful connection to consumers. It's why Google Shopping could never make a significant dent. That it's an intangible search & ad company, that essentially no consumers buy anything from (in the Amazon sense), will persistently haunt them as the Internet of things / AI / robotics / autonomous whatever continues to rapidly move forward in the consumer space. Maybe they'll start to change that with Pixel.
Meanwhile, Amazon is increasingly touching consumers at every possible turn, including soon to be probably hundreds of physical stores of various types. It's like trying to compete with Tesla in autonomous, doing everything just in the lab, without having any physical driving data.
If you think of Google's top services that touch consumers: search, gmail, maps, play store, youtube - it's all a million miles away from the tangibility that Amazon has in physically touching consumers on a frequent basis. Google is completely out of its element when it comes to doing the kind of consumer work that Amazon does, including when it comes to customer support. Today Amazon can physically get into your home in several ways, Google will probably always struggle with that (a fundamental difference in how the two companies are oriented and wired).
> I didn't even know till now that Google has a new personal assistant your thing that's not Google Now.
I'm guessing you're an iPhone user and therefore don't have much interest in Android...? There were a lot of positive reviews for the Pixel, both on HN and elsewhere, and most of them seemed to cite the new assistant as the main differentiator between the Pixel and other phones.
You're probably right that Google doesn't do well with marketing in general but it's hard to believe that someone who follows tech news wasn't aware of their assistant before now.
As someone who reads HN most days and is fairly interested in Android, I'm aware that Pixel and Google Home are somehow different from Google Now but I have no idea what that difference is supposed to be.
I keep hoping back and forth between iPhone and Android. However, I didn't need a phone when the Pixel came out and thus didn't pay much attention to it beyond noting that it's more premium and thus its successor might be a real alternative for my next phone.
That being said I also wasn't very interested in the iPhone 7 for the exact same reasons. However, I heard all about it organically at work because people keep talking about the cameras, missing headphone jack etc. Same with Alexa where co-workers talk about ordering stuff through it via prime with visitors on the house just to show off the device to them.
I agree that Google Assistant is a strong differentiator for marketing the Pixel phone, but, I think Google is making a mistake not pushing it to more Android phones. For that matter, perhaps they should try to make it available as an app on iPhone and iPad ASAP also. I find the Google App to be very usable on my iPad, so why not update that same app to support Google Assistant?
If Amazon isn't a search company, then Google isn't an OS company. Search is pretty integral to Amazon, search being the main way that anyone finds anything on their site. Sure, selling other company's products is what they're known for, but they do a lot more than that. And funny enough, a search engine is one of them. A9 (an Amazon company) develops custom search engines and search technologies.
You can't pigeonhole conglomerates like that. Google is an email company, Amazon is a cloud hosting provider, Microsoft is a tablet PC maker, Samsung is a heavy equipment company. No wait, Google is a browser vendor, Amazon is a bookstore, Microsoft is the maker of an office suite, and Samsung makes hardware for the iPhone.
Or maybe they're all tech companies and specialize in multiple aspects of that industry.
Actually, it's pretty clear that Amazon is not a search company - nor is Google. Amazon's market is digital retail, while Google's is digital marketing.
Amazon's tech is all an effort to extend their reach in their primary business: selling products (both physical and digital) to customers. Google's tech, on the other hand, is all in an effort to extend the reach of their ad and marketing services.
Amazon figured out a while ago its tech can extend their primary business and also be sold to other businesses on the side (AWS and 3p sellers/FBA). This is a broadening that google hasn't been successful at yet.
He's deriving what a company is based on their revenue streams. What he forgets is that without Google being as good as it is at search there wouldn't be that lucrative revenue.
Amazon actually had a general purpose search engine once. That's what a9.com was. The company A9.com [1] is still around as an Amazon subsidiary [2], the search engine is not.
A grand total of 0 of the things I've tried to ask my Alexa has been search. If I want to search, I pick up any of half a dozen devices scattered around the room that has a screen, because I don't want to have to wait to have results read to me.
What appealed to me about it was that the examples given were simple and practical.
The other part of assistance is integration with the physical realm.
As a global store front and supply chain, Amazon has a serious advantage in that area... manufacturers are more inclined to cooperate and integrate with the Alexa ecosystem. Amazon's end-game is obviously "sell more stuff" - obvious benefits for the manufacturers. I can't see a clear vision behind Google's plan.
However, so far every device is a walled garden, a locked-in ecosystem - bleh, huge turnoff.
Is there prior work in developing an an open assistant/querying protocol for these kinds of devices? Some standard way for devices to get data after the device has handled wakewords and speech-to-text? It's a fascinating domain, but I don't have enough knowledge to bootstrap myself yet.
I would put forth that Google isn't a search company anymore. They spend more resources filtering out all the junk on the web. Search is a secondary function for them. The end-user doesn't see the filtering; publishers and merchants trying to jockey for position do.
Wow, we've come a long way from the outrage over the Kinect supposedly "quietly listening all the time" to these products now being the next great thing.
Unfortunately the maxim continues to prove true: "nobody cares about privacy or security."
I've come to the conclusion that technical loss of privacy is inevitable because customers don't care, and that when the you know what does eventually hit the fan the problem will have to be solved with regulation. There will be HIPAA/PCI type laws around home audio, location data, etc., and a renegotiation of wiretap and search and seizure laws. There may also be some legal limits on private data collection and retention.
Of course malicious hackers and secret parts of governments won't obey these laws. We are in some uncharted waters.
However the strength of something like Amazon's devices is the ecosystem and the microphones that they ship with, which are expensive to buy separately.
I think, as proven by Glass, that the public seems to view Google as nice but creepy. Whereas Amazon is usually tops in customer brand/satisfaction type of scores. My guess is that this is why even though Google definitely has a technologically superior product (better understanding, better inference, etc) they seem to be losing the "hearts and minds" of consumers to Amazon's more bare MVP technology.
Amazon are Borging retail, and have a major problem with Bezos being widely viewed as an asshole, with strong justification.
Mind: much the rest of the tech world does little better. Microsoft, Oracle, Apple, Facebook, IBM, and more.
I'm inclined to think that its the sector itself which selects for this, though in the selection process, the personnel (and especially management) are very mich part of the problem.
Whats to prevent someone from connecting this to their smart door locks? Then if that person has a big party, and someone with bad intentions shows up, they can simply say "Alexa, at 3:30am tomorrow, turn off the alarm system, and unlock the doors". Or "Alexa, turn off the dialysis machine".
Sure Amazon might make it more secure in the future, but probably only after enough of these unintended consequences stack up.
I also understand that the average HN reader is probably not rushing out to introduce a single point of failure into their home security system. My point is that Alexa is essentially saying "don't worry about how it happens, just dream big about what awesome stuff might happen".