I like the fact the ad managed to show the potential unwanted side affect of such device in a manner that everyone understand. I do see the appeal of smart devices and it has to start somewhere and all such start would be painful. But with the issues of IoT, the rampant ignoring of privacy by big corp and govt. alike, the lack of understanding of potential impact by even the knowledgeable, the apathy of the masses makes it a very dangerous time for some of these innovations.
My kid wanted a fitness watch for Christmas. One morning I received bunch of emails about purchases on Amazon, but I was too busy to look and I knew that it was probably my wife Christmas shopping. In the evening I had a chance to look at the emails - 6 orders of watches $175 each. I had time to cancel all but one. So my kid got black watch for Christmas, even though he wanted blue. At least size was correct. Amazon should not have switched Alexa purchasing on by default.
The device really should implement speaker identification before accepting commands, yeah. Seems like a task a machine learning focused company like Google really should be able to accomplish.
And require enough entropy to make it hard to guess. Maybe now is the time to start learning Khoisan to get some unusual phonemes in your pronunciation toolkit.
But there is a reason they have the wake names they have - they're unique and don't sound like other phrases. I recall reading something about how Alexa was chosen because "ah-LECK-sah" is a very unique sound. I strongly suspect this is why we have "oh-kay Google" and not just "Google".
So if people have to choose their own, they'll have to choose one that is also not easily mistaken. Something tells me that would be a very annoying process.
I have friends who have an alexia. It wakes up to unrelated words on its own all the time. How can a company like google or amazon know what words are commonly used in a household anyway? protip: they can't.
It's not that they chose those keywords for uniqueness, they chose them for branding purposes.
I'd imagine we don't have just "Google" because the word Google comes up in conversation in many contexts that are not the user requesting assistance from Google Home.
Motorola Android phones like Moto X Pure/Style had voice identification. You had to record the wake phrase 3 times and the voice assistant would only listen to your commands.
It worked quite well, I don't know why Google Home doesn't have anything like this.
No company will give up an information vacuum in a busy living space. And if a company did they would just get acquired by a company that would change the software and TOS to become an information vacuum.
It's sad that Google just deactivated the phrase. There's so much potential in playing with device-originated communications:
- Siri and Google could work together a script to detect each other (usually setup by users who want fun) and start an actual discussion, perhaps from a movie, a comedy, etc. Bam! Movie advertising, and the surprise effect makes people share it virally.
- TV ads-triggered events could generate a health discussion next time the person goes to Burger King, or detect friends who were triggered at exactly the same time and suggest to start a convo about this particular TV show. Would be fun if watching TV became a social experience thanks to Android or Alexa.
At that point, instead of being triggered and filtering out wrong occurrences, they could build something upon it.
Or maybe because they planned to expand the campaign globally (as they often do with their ad campaigns), and Android happens to be by far more successful globally?
The Norwegian submission to the Montreux comedy festival in 1983 [1] had as a plot point how one of the characters got locked out of his flat because he had a cold that changed his voice enough that it wasn't recognised by his building security system. It'll take a long time before we have all of the kinks sorted out...
It seems to me Google could easily mitigate against this by fingerprinting the query. If anomalously high levels of the same fingerprint appear at the same time, ignore it.
Google did find some clever way to stop it working. The article says within hours of the commercial being reported on, it had stopped activating the device.
Easy to make an exception for that particular pattern but that doesn't stop someone else from doing this again.
The more successful a medium is the bigger the chance that it will be abused for the equivalent of spam. Burger King is acting like those two lawyers that wrote a how to spam guide ('How to make a fortune on the information superhighway').
Don't know what you mean by "that" in your comment, but fingerprinting all requests (which they probably already do) and then looking for elevated rates is surely a generalisable pattern?
Actually I think this is quite effective at stopping companies from doing this in the future. Why dump millions of dollars into creating an ad and paying for TV airtime when Google can just shut you down within hours?
I don't live in the US and I never saw an ad for Alexa or G Home - and I have a problem understanding how can they be useful beyond a fancy gimmick. Can you seriously trust it with your pizza order even if you don't know what the menu is and what you want? Then how many times a week do you order a pizza? Maybe I it is just me being old, please educate me.
Any pizza place in a city where it’s common that a school or a professor would order hundreds of pizzas.
Our high school, due to having no cafeteria for a while, ordered once or twice a week hundreds of pizzas from several pizza places.
And we have a professor at our university who always promises to give everyone free pizza if more than 25% of the class are still there the week before the exam. Sometimes it works, sometimes too many fail.
But ordering many pizzas, especially online, is possible.
Although some places will call you back to confirm
In college I worked a pizza joint. 300 pizza orders were not uncommon, but anything over 200 dollars got a callback to make sure it was legit. But hey, if the card clears, noone cares.
Easily solved, though, by adding additional confirmation steps if the amounts are above a certain level or the order deviates from your usual patterns. I look forward to giving Star Trek type authorization codes when ordering pizzas...
A friend of mine has an Alexa, and I never would have guessed how much fun it is to be able to say "Alexa, play <some song>" and have it just happen. The marketers far overstate its significance, like marketers do, but it is a handy little gizmo to have around.
So far, the best combination of neat and useful for me has been integration with Logitech Harmony and having Alexa start turning stuff on while I approach the couch. It feels so futuristic to be able to shout in to the air that I would like to watch some TV or play a game and have all that stuff just come up for me. If I spent more time with it, I could probably even get it turning to specific channels or launching Rocket League automatically. I tried doing the normal shopping list routine with it, or asking it to set timers, but none of that has really stuck in the same way that "Alexa, turn on AppleTV has." I do like her sense of humor though, and "tell me a joke" is a mainstay.
I like to use it for situationally-appropriate accompaniment. Just won a game of Pandemic? "Alexa, play 'We Are the Champions'." Just lost a game of Pandemic? "Alexa, play 'It's the End of the World as We Know It'." :)
My fiance and I have just finished serving up a plate of food for dinner. We go and sit down on the couch and want to start eating, but someone has to futz around with their phone for a couple of minutes to get an episode of TV to play while their food gets cold.
After getting a home one of us can just say "Ok google, play The Next Generation from Netflix on the TV" while walking towards the couch. It plays the next episode right were we left off yesterday.
One of us gets up to go to the toilet, normally it means screwing around with the phone, and oh look, the chromecast disconnected from the phone so we have to reconnect again. It's easy to spend a whole minute just to get the damn episode to pause.
After getting a home it's just a matter of saying "Ok google, pause the TV"
One of us gets up to go to the toilet, normally it means screwing around with the phone, and oh look, the chromecast disconnected from the phone so we have to reconnect again. It's easy to spend a whole minute just to get the damn episode to pause.
Reality: if you have a slightly non-standard accent, you will be shouting at your device five times before anything happens. While I could just have pressed the play button on my Apple TV remote and be done with it (no fiddling with a phone necessary).
Then there are all kinds of cultural realities that are apparently not considered in Silicon Valley. My favorite five pet peeves:
(1) Our family name is 'de Kok'. Very often these assistants totally mess up when I ask e.g. to call family (either they start complaining about dirty language or literally search for 'cock').
(2) We are a multilingual family, languages used in our household are German, Dutch, and sometimes English. None of these voice assistants work with multiple languages. So, the only alternative is to choose one language and stick with it. Unfortunately, if you choose English you are having a hard time navigating in Germany, looking up a German TV program, or calling a Dutch or German friend. If you set the language to German, you forgo looking up shows on Netflix by their English titles reliably.
(3) Usability goes down the drain once you have kids. Voice assistants typically do not work well when multiple people are speaking at the same time. If you are lucky, they are just confused, if you are unlucky they start doing whatever they infer from a semi-random mix of words. You cannot tell your three year old to be quiet, because mom or dad has to tell Siri what to do. (I use Siri sometimes, and our daughter currently likes to scream 'hey siri' repeatedly when I use Siri.)
(4) Voice assistants are unusable in most environments where there are other people.
(5) Voice assistants, especially in-house voice assistants are a potential privacy violation. So, you have to ask people that visit you if they are OK with your conversations being recorded (accidentally or intentionally) and switch off the device if they do mind the latent spying.
tl;dr: voice assistants only work well in monolingual families where all of the family members have a perfect accent and where none of your acquaintances mind the privacy violations.
> voice assistants only work well in monolingual families where all of the family members have a perfect accent and where none of your acquaintances mind the privacy violations.
I had hard time getting Google to understand my speech 2 years ago. These days the error rate has come down a lot. From 60-70% to around 95-96%. For me Google home won't be worth it till it reaches more than 99% accuracy but as more people use it the accuracy increases
> Reality: if you have a slightly non-standard accent, you will be shouting at your device five times before anything happens.
I don't know about Google Home or Siri, but my Alexa easily handles my broad Norwegian accent the vast majority of the time.
There are a few things (e.g. weird album titles) that are more likely to cause problems, but in my experience the things that it has problems with are rarely accent related as when something is problematic my son who does not have a comparable accent runs into the same misrecognition.
(EDIT: I have no doubt that there are accents it will have problems with - my point is that you certainly don't need to have perfect pronunciation for this any more)
> While I could just have pressed the play button on my Apple TV remote and be done with it (no fiddling with a phone necessary).
That requires finding the remote first. You may be organized enough for that never to be an issue. I've given up on the normal remote and rely exclusively on apps now, but that also involves grabbing my phone or tablet and a couple extra steps.
> (4) Voice assistants are unusable in most environments where there are other people.
That's fine. They don't need to be usable everywhere. I have no interest in using one on my phone in public for example. But I like having one in my living room, and I rarely have enough people around that it's a problem. In my experience with it so far, people have already learned to hush down automatically when they hear you give the keyword (but I do agree that won't work with a lot of people present - it works with a few people around).
> "tl;dr: voice assistants only work well in monolingual families where all of the family members have a perfect accent and where none of your acquaintances mind the privacy violations."
Indeed, and "perfect accent" varies from device to device. My wife and I are both from the Atlanta, Georgia area, yet we have distinctive accents; hers is much more "Southern" than mine as I grew up in several different states across the Southeast and Southwest, whereas she lived in the same house from birth to college, then moved back after graduating.
We've found that her Nexus 6 understands me better than her even though she talks to it far more often than I do, and the reverse is true for our Shield TV despite it using the same backend as the phone. She has to try two or three times to get her phone to understand a command, where I can get it right the first time. On the Shield, I will tell YouTube to look for "funny cat videos" and I get "funny hat videos" every damn time, while she gets cats.
Dude, I can ask my phone to show me information about Vishwaroopam and then navigate to Bermondsey station and then search for Taqueria Guadalajara right after. I literally just did it.
At some point you're going to have to face a fact: you're an irrelevant minority. That doesn't mean that the tool only supports a narrow use case. In my case, I found a way to my Tube station in London, information about a Mexican restaurant in San Francisco close to my home there, and then found information about a Tamil movie. First try each time.
You made a prediction that was totally and completely wrong based on a single data point. The prediction was poor quality, proven by the fact that I can do this. I don't want my HN cluttered with low quality predictions based on low quality data. So I voted you down.
Thanks god Alexa and G Home solves all of our first word problems caused by our laziness and digital addiction. There were times I thought the Cyberpunk motto "High Tech, Low Life" would be just a joke.
Maybe you should add a rock of salt, or two. I read the GP's description as sarcasm. Even an iPad offers a better experience than what they describe. Or they're on Android.
Still, the cyberpunk motto is bound to become true.
Back in my day, we had a device that had a dedicated button for 'pause'. It worked using infrared, a simple invisible light system that doesn't need pairing or a wifi connection. The device (called a "remote control") could also run for years on a single AA battery, using little to no power when no buttons were pressed.
The downside was that it had a lot of buttons, 90% of which you don't use. It also tended to get lost a lot. And we didn't have Netflix, so either we'd have to use video tapes (which was cool, insert and hit play), or DVDs (which were a bit less usable, they usually had unskippable warnings, trailers, and some weird menu you had to get through before playing the movie / TV show). Or just watch what was on TV at the time, which worked just fine for us because we'd eat and go to bed at set times.
My tl;dr point is that things just used to be simpler and more usable. I don't like having to use an app to start stuff. I don't like the four remotes I have lying around atm either (TV, set-top box for digital television, one for the soundbar (idk why it's that big), and either a controller for the ps4 or for the apple TV. The latter two kinda fulfill the same role though, as a netflix machine.
I have no interest in these smart home devices and I'm not a Luddite. These things just complicate life. Just like my fancy new dishwasher which thinks it's a good idea to beep constantly once my dishes finish washing at 1am.
If it complicates your life, it's a bad device. My Echo does not do that, but that's probably because I only have a limited number of things I really want it to do, and it does them just fine.
My sodding washing machine does this after completing a washing programme and there's no way to disable it. Seriously, does the manufacturer think I'm going to completely forget I needed clean clothes and bedding for the week?
I also notice my new microwave sits there and keeps chirping at me after heating food until I open the door, it's not as loud and as insistent as the washing machine, but still totally unnecessary behaviour.
A more sensitive soul could eventually be driven bonkers if every flippin' thing in your house and car beeped and booped at you every minute of the day.
I always buy appliances that are mid-to-lower range. There's none of the over engineered smart technology and they last decades rather than years.
I think once you've spent enough time submerged in technology, you start to appreciate the simpler things in life. A smart washing machine or microwave isn't one of them.
I didn't buy these appliances (rental), my last apartment had basic analog appliances. They're so much better. I really do not need digital inputs on my appliances. I definitely do not need WiFi or smartphone control!
I have two remotes: One to turn the TV on and off or change the volume, and another, much smaller one to control my Android box running KODI. I just re-used the Amazon fire TV stick remote. Works great! 7 buttons.
Simple and usable is still possible, but you have to do your homework.
If you want to get that down to "one" check out Sideclicks[1]. A small universal remote that attaches to the fire TV stick remote. Not associated, just a user that appreciates not having to keep track of multiple remotes.
Most devices that aren't Chromecast have a remote. (Roku, Apple TV, for example). I have no idea whether those would work with Google Home since I don't have one. I have a Chromecast and Echo Dot (Alexa) and they don't work together (nor Alexa and Apple TV). I'm guessing Amazon wants me to have a Fire device for TV and phone and tablet. For that matter, first party Apple apps on my phone (Apple Music) don't work with Chromecast, only AirPlay (to Apple TV). Of course YouTube on the phone, being a Google product, will only work with Chromecast. I can ask Alexa to stream from my Audible or Amazon Music Unlimited accounts (when I had it temporarily), but not Apple Music (and I assume not Google's music service). It sucks that all these services mostly refuse to work with each other. There is some overlap, such as YouTube app on Apple TV, but using the YouTube app on the phone is a lot more usable (even with their confusing UI).
I should probably stick with free software and non-DRM content.
Having said a that, I do use the Echo Dot to do hands free timers (like when I'm cooking), calculations in general and for cooking ("Alexa how many tablespoons are in a quarter cup"), look up definitions of words (while I'm sitting on the couch reading a book), and my kids ask it how to spell things. (The only reason I have one is I got it for free at AWS re:Invent conference.)
A small technical point here, but with HDMI CEC you can use your regular remote for pausing chromecasted stuff. I use the remote typically for pause / play / volume, and my phone to pick what to watch (also the phone app has a handy "go back 30s" button).
For me, these smart hubs won't be particularly useful until I can combine a bunch of functionality into one thing. I have the heating set to come on in the morning, but our morning schedule is irregular so I have to adjust it most evenings to come on a bit before we wake. Then I set the lights to come on, as I find it easier to get up when it's light, and set an alarm. What I want is just to say "wake me up at 5:30 tomorrow" and have all that just work.
The individual things (alarm, light and heating) are all individually beneficial though, despite the evenings faff. This is also only a minor problem too, these are not daily herculean tasks I barely make it through, but it's still a level of hassle I'd like to smooth over.
I have a PC plugged on my TV. I can browse local or cloud (Netflix, Youtube, TED, whatever) collections with the TV remote, get things to play, stop, whatever. When I want to do something more complex that would be too time consuming with a remote (not common), there's a wireless keyboard that stays on a drawer.
You can still make things simple and easy. What we are seeing is a complete lack of trying. Probably because those companies just love to have an always listening microphone inside your home, 1984 style.
Certainly there's nothing an Echo or whatever can do that you can't do with any properly-set-up PC or phone. The Echo just does it faster and hands-free. It's not life-changing or revolutionary, but it is handy.
I love mine(s) now. I made a bridge between my RF wall switches [0], outlets and so on and I can turn things on/off using Alexa.
The best is at night before going to bed, I can say 'Alexa, turn off downstairs' and that will turn off all lights, lamps, speakers, amplifiers.
But yes, I can also tell her to play some music thru the speakers, and 'Alexa, turn on speakers' and the music automagically starts playing on the stereo.
Then there's all the other bits, but the timers we use really all the time, the 'Alexa, what's up' in the morning is pretty nice too...
I have alexa and we use it CONSTANTLY as a timer for our kid. You have 2 minutes in timeout. Boom alexa does it. I can take him to timeout and get that started at the same time.
It also makes it so that I'm not saying when he gets out but a device is saying when he gets out. He doesn't tend to argue with Alexa but he will argue (negotiate) with me so it helps in that way.
I also use it to play music and read books to me. In the future I'm going to set it up to work with my smart home, but I haven't set that up because we're selling our house. Was it worth the price point? No, but now that I have it it's become part of my daily life.
That's quite a brilliant use case. This is why I believe chat bots and by proxy voice assistants like this will be standard technology, despite the risks (having an always on microphone is not entirely necessary). There are probably thousands of these niche use cases that we haven't thought of because these are new interface form factors.
Tons of opportunity for innovation/products here. The only downside (besides the need for server voice processing) is that you are building a company on a platform. The whole Twitter and Facebook platform debacles showed the risks of trusting your business life support to a bigger indifferent company - who may just steal your idea and implement it.
Longer code-phrases actually give the voice recognition system lower probability of false positives and false negatives. Additionally, the code phrase needs to be something slightly obscure right now; you don't want the machine triggering every time you walk through the front door and yell "Honey, I'm home! How was your day?"
I think this is totally fine by Google. "O.K. Google" would be a trade or service mark of Google, and you'd need to obtain permission in order to use this mark in an advertisement.
The way the ad depicts BK and Google "working together", to speak a favourable description of a BK product, suggests some sort of Google endorsement for BK.
Without that endorsement being approved, it's right that Google would act to stop the ad from suggesting or using an endorsement.
From my point of view, this is pretty risky for BK to do so. And either they didn't understand the risks, or they knew the repercussion would not be expensive for them, or they just decided to take the risk.
To give BK the benefit of the doubt from a strategic perspective, maybe the thinking was, "We know Google will react negatively or shut it down, but that reaction will also get coverage, and will promote talk of our ad even further. Thanks Google!" Still, I'm not marketing person, but that type of thinking, seems to me sort of like "poke the bear" for benefit, and seems quite bizarre to me. I feel it's strange if that is really how marketing people calculate.
The interesting question is if "Google" is even still a trademark of Google.
Considering many dictionaries list it as "to google: to search in the internet with a search engine, e.g. Google® or Bing®", one could argue it has become generic, and the invocation isn’t actually "O.K. Google: ", but "OK, google: ", using it as imperative form of a verb.
That said, this argument is today very silly, as it relies on the letter-of-the-law interpretation, and wouldn’t fly in most courts yet. In a few years, with "to google" being used more commonly, it might actually.
> "O.K. Google" would be a trade or service mark of Google
It is an utilitarian mark. You can't trademark those. Or better, you can, you just don't get protections against utilitarian uses, like other manufacturers using the same mark for compatibility or companies using it to access the service.
I don't know much about this, but when I set Siri up I think it does some voice recognition so that I am the only one who can control the phone.
Couldn't the same thing be done with Google and Alexa?
During installation, one could add his own voice, and then your command "Google, add this voice" would allow you to add another member of the household.
Does that work for the hands-free activation? The voice lock on Siri doesn't apply to queries once it's activated, only to the activation itself.
E.g. if I say "hey Siri" while my phone has its screen off, it wakes up and waits for a query. If my girlfriend (or anyone else) says a query at this point it will respond to it. But if the screen is off and my girlfriend says "hey Siri", nothing happens.
I don't even need to speak for that to happen. I just bend over with the phone in my back pocket and it wakes up, beeps, waits for me to say something (I say nothing,) boops, then goes "Sorry, I didn't get that."
This is a clever hack, and arguably appeals to a core Burger King demographic. Whether they experienced it themselves, or just read about it. Because it'll get shared and discussed. And the next time you want a burger, maybe ...
What's evil, I believe, is that ultrasonic cross-device tracking technology. This was just cute.
I think it's a great move to show the general public how insanely insecure and intrusive the IoT and always-listening cloud devices can be. People brought it to light with google and with amazon - nothing changed and unaware people lapped up the devices, hopefully a demonstration thats more tangible to the general public phones home (pun intended).
Still waiting for a shock jock to say "ok google, show me pictures of naked kids" during the morning rush. People will die scrambling for their phones, but it will finally be the end of this silliness.
There are radio ads running in Australia or New Zealand (I forget which) that instruct Siri to turn airplane mode on. They are PSAs about texting or dicking with your phone while driving, so I guess they're helpful, but still.
I don't understand why these Telescreen ( ok, teleblock ) technologies don't require voice print access control as default. Voice biometric is older than this cloud voice NLP tech. Also, it's amazing to me that human psychology and the zeigeist requires everyone flip out over Google Glass, but always on microphones linked to the Company, totally fine. Tolerated, in fact, behind the point of inconvenience. The other incongruous observation I want to add is voice is taking off in the home, but people using voice typing and control in public isn't as accepted. There's an essay in this somewhere.
Don't visit any of your friends who have bought it.
Don't go out in public.
Don't be a part of society.
Don't even SMS or email anyone since they might have one too.
The fight to reduce surveillance must happen even when the people who are fighting for privacy don't purchase the products that everyone else buys. If 99% of people have bought Google Home in the future, what good does it do to tell someone "hurr durr don't buy it?" No good at all.
This episode is a good learning experience for these people about the dangers of having an all listening microphone in your house. The best thing we can do, is keep educating people.
The one thing we can do is create something like UL/Consumer reports for privacy. It is desperately needed. People automatically check this site to see what the exposure is of having a certain device in your home. I hope this comes to pass.
A computer with a microphone and a cell phone both have the social contract of only "on demand" recording, as in, the user directly turns on the microphone and is aware that the recording is happening.
A trojan horse like Alexa and Google's device sneakily reverses this contract to the benefit of the corporation/government. People still feel the listening devices are "on demand" but in fact are being listened to all the time.
Education is really the only way I can see that any of us at the ground level can effect any change, here.
Firstly, if Google et all, allowed us as users to customise the name of our assistant. That would be appreciated by all users. Alas they do not and no sign they will either anytime soon.
Second aspect, they have no voice recognition tied in to these assistants and that would curtail all this. Though not ideal it would be a high improvement from what we have currently.
But until we have customisable names for our assistants, voice recognition, then this and much more will carry on happening.
Do you have a landline and screen calls, well somebody could ring you up and speak to your assistant currently, that is an issue still.
For me I don't see things getting any better until we have the above tied into some additional user biometrics, be that array microphone tracking that is checked with visual recognition of some form.
Though a cheap simple solution would be a watch with one button that you pressed and spoke into and that was linked to the assistant. That would not be hard and work much better.
131 comments
[ 2.0 ms ] story [ 203 ms ] threadNo.
It's Brazil.
https://en.wikipedia.org/wiki/Brazil_(1985_film)
(Agreed it's all becoming a bit surreal)
http://www.theverge.com/2017/1/7/14200210/amazon-alexa-tech-...
[0] https://support.google.com/nexus/answer/6093922 - last entry.
Just allow every buyer to give the device their own unique wake name.
So if people have to choose their own, they'll have to choose one that is also not easily mistaken. Something tells me that would be a very annoying process.
It's not that they chose those keywords for uniqueness, they chose them for branding purposes.
I watched Sneakers recently and I think it really holds up.
It worked quite well, I don't know why Google Home doesn't have anything like this.
By the time this shit becomes useful, it will be able to recognize my voice, like any good servant. And it will not be tethered to the cloud.
It's sad that Google just deactivated the phrase. There's so much potential in playing with device-originated communications:
- Siri and Google could work together a script to detect each other (usually setup by users who want fun) and start an actual discussion, perhaps from a movie, a comedy, etc. Bam! Movie advertising, and the surprise effect makes people share it virally.
- TV ads-triggered events could generate a health discussion next time the person goes to Burger King, or detect friends who were triggered at exactly the same time and suggest to start a convo about this particular TV show. Would be fun if watching TV became a social experience thanks to Android or Alexa.
At that point, instead of being triggered and filtering out wrong occurrences, they could build something upon it.
Or maybe because they planned to expand the campaign globally (as they often do with their ad campaigns), and Android happens to be by far more successful globally?
[1] Happy New Century: https://tv.nrk.no/serie/montreux/fuha00005782/30-04-1983 (in Norwegian only; it's set on New Years Eve 1999 - it's quite dated)
Well done Burger King, after all -- what good is Google if you can't stand on their shoulders, right?
The more successful a medium is the bigger the chance that it will be abused for the equivalent of spam. Burger King is acting like those two lawyers that wrote a how to spam guide ('How to make a fortune on the information superhighway').
Our high school, due to having no cafeteria for a while, ordered once or twice a week hundreds of pizzas from several pizza places.
And we have a professor at our university who always promises to give everyone free pizza if more than 25% of the class are still there the week before the exam. Sometimes it works, sometimes too many fail.
But ordering many pizzas, especially online, is possible.
Although some places will call you back to confirm
"Alexa, play some soft rock"
"Alexa, set a timer for 10 minutes"
"Alexa, skip this track"
"Alexa, how long do I boil an egg for"
After getting a home one of us can just say "Ok google, play The Next Generation from Netflix on the TV" while walking towards the couch. It plays the next episode right were we left off yesterday.
One of us gets up to go to the toilet, normally it means screwing around with the phone, and oh look, the chromecast disconnected from the phone so we have to reconnect again. It's easy to spend a whole minute just to get the damn episode to pause.
After getting a home it's just a matter of saying "Ok google, pause the TV"
Reality: if you have a slightly non-standard accent, you will be shouting at your device five times before anything happens. While I could just have pressed the play button on my Apple TV remote and be done with it (no fiddling with a phone necessary).
Then there are all kinds of cultural realities that are apparently not considered in Silicon Valley. My favorite five pet peeves:
(1) Our family name is 'de Kok'. Very often these assistants totally mess up when I ask e.g. to call family (either they start complaining about dirty language or literally search for 'cock').
(2) We are a multilingual family, languages used in our household are German, Dutch, and sometimes English. None of these voice assistants work with multiple languages. So, the only alternative is to choose one language and stick with it. Unfortunately, if you choose English you are having a hard time navigating in Germany, looking up a German TV program, or calling a Dutch or German friend. If you set the language to German, you forgo looking up shows on Netflix by their English titles reliably.
(3) Usability goes down the drain once you have kids. Voice assistants typically do not work well when multiple people are speaking at the same time. If you are lucky, they are just confused, if you are unlucky they start doing whatever they infer from a semi-random mix of words. You cannot tell your three year old to be quiet, because mom or dad has to tell Siri what to do. (I use Siri sometimes, and our daughter currently likes to scream 'hey siri' repeatedly when I use Siri.)
(4) Voice assistants are unusable in most environments where there are other people.
(5) Voice assistants, especially in-house voice assistants are a potential privacy violation. So, you have to ask people that visit you if they are OK with your conversations being recorded (accidentally or intentionally) and switch off the device if they do mind the latent spying.
tl;dr: voice assistants only work well in monolingual families where all of the family members have a perfect accent and where none of your acquaintances mind the privacy violations.
Edit: added some more annoyances.
Edit 2: heavy downvoting :/. Please explain why!
So, bachelor Silicon Valley engineers.
I don't know about Google Home or Siri, but my Alexa easily handles my broad Norwegian accent the vast majority of the time.
There are a few things (e.g. weird album titles) that are more likely to cause problems, but in my experience the things that it has problems with are rarely accent related as when something is problematic my son who does not have a comparable accent runs into the same misrecognition.
(EDIT: I have no doubt that there are accents it will have problems with - my point is that you certainly don't need to have perfect pronunciation for this any more)
> While I could just have pressed the play button on my Apple TV remote and be done with it (no fiddling with a phone necessary).
That requires finding the remote first. You may be organized enough for that never to be an issue. I've given up on the normal remote and rely exclusively on apps now, but that also involves grabbing my phone or tablet and a couple extra steps.
> (4) Voice assistants are unusable in most environments where there are other people.
That's fine. They don't need to be usable everywhere. I have no interest in using one on my phone in public for example. But I like having one in my living room, and I rarely have enough people around that it's a problem. In my experience with it so far, people have already learned to hush down automatically when they hear you give the keyword (but I do agree that won't work with a lot of people present - it works with a few people around).
My TV is so old the remote is still wired to it. Finding the remote is never an issue.
Indeed, and "perfect accent" varies from device to device. My wife and I are both from the Atlanta, Georgia area, yet we have distinctive accents; hers is much more "Southern" than mine as I grew up in several different states across the Southeast and Southwest, whereas she lived in the same house from birth to college, then moved back after graduating.
We've found that her Nexus 6 understands me better than her even though she talks to it far more often than I do, and the reverse is true for our Shield TV despite it using the same backend as the phone. She has to try two or three times to get her phone to understand a command, where I can get it right the first time. On the Shield, I will tell YouTube to look for "funny cat videos" and I get "funny hat videos" every damn time, while she gets cats.
At some point you're going to have to face a fact: you're an irrelevant minority. That doesn't mean that the tool only supports a narrow use case. In my case, I found a way to my Tube station in London, information about a Mexican restaurant in San Francisco close to my home there, and then found information about a Tamil movie. First try each time.
You made a prediction that was totally and completely wrong based on a single data point. The prediction was poor quality, proven by the fact that I can do this. I don't want my HN cluttered with low quality predictions based on low quality data. So I voted you down.
https://www.youtube.com/watch?v=HbDnxzrbxn4
Still, the cyberpunk motto is bound to become true.
The downside was that it had a lot of buttons, 90% of which you don't use. It also tended to get lost a lot. And we didn't have Netflix, so either we'd have to use video tapes (which was cool, insert and hit play), or DVDs (which were a bit less usable, they usually had unskippable warnings, trailers, and some weird menu you had to get through before playing the movie / TV show). Or just watch what was on TV at the time, which worked just fine for us because we'd eat and go to bed at set times.
My tl;dr point is that things just used to be simpler and more usable. I don't like having to use an app to start stuff. I don't like the four remotes I have lying around atm either (TV, set-top box for digital television, one for the soundbar (idk why it's that big), and either a controller for the ps4 or for the apple TV. The latter two kinda fulfill the same role though, as a netflix machine.
I also notice my new microwave sits there and keeps chirping at me after heating food until I open the door, it's not as loud and as insistent as the washing machine, but still totally unnecessary behaviour.
A more sensitive soul could eventually be driven bonkers if every flippin' thing in your house and car beeped and booped at you every minute of the day.
I think once you've spent enough time submerged in technology, you start to appreciate the simpler things in life. A smart washing machine or microwave isn't one of them.
Simple and usable is still possible, but you have to do your homework.
[1] = https://www.sideclickremotes.com
I should probably stick with free software and non-DRM content.
Having said a that, I do use the Echo Dot to do hands free timers (like when I'm cooking), calculations in general and for cooking ("Alexa how many tablespoons are in a quarter cup"), look up definitions of words (while I'm sitting on the couch reading a book), and my kids ask it how to spell things. (The only reason I have one is I got it for free at AWS re:Invent conference.)
For me, these smart hubs won't be particularly useful until I can combine a bunch of functionality into one thing. I have the heating set to come on in the morning, but our morning schedule is irregular so I have to adjust it most evenings to come on a bit before we wake. Then I set the lights to come on, as I find it easier to get up when it's light, and set an alarm. What I want is just to say "wake me up at 5:30 tomorrow" and have all that just work.
The individual things (alarm, light and heating) are all individually beneficial though, despite the evenings faff. This is also only a minor problem too, these are not daily herculean tasks I barely make it through, but it's still a level of hassle I'd like to smooth over.
You can still make things simple and easy. What we are seeing is a complete lack of trying. Probably because those companies just love to have an always listening microphone inside your home, 1984 style.
The smart-tv is almost as bad, I'm just going to connect an actual computer to the TV and be done.
The best is at night before going to bed, I can say 'Alexa, turn off downstairs' and that will turn off all lights, lamps, speakers, amplifiers.
But yes, I can also tell her to play some music thru the speakers, and 'Alexa, turn on speakers' and the music automagically starts playing on the stereo.
Then there's all the other bits, but the timers we use really all the time, the 'Alexa, what's up' in the morning is pretty nice too...
[0]: https://github.com/buserror/rf_bridge
It also makes it so that I'm not saying when he gets out but a device is saying when he gets out. He doesn't tend to argue with Alexa but he will argue (negotiate) with me so it helps in that way.
I also use it to play music and read books to me. In the future I'm going to set it up to work with my smart home, but I haven't set that up because we're selling our house. Was it worth the price point? No, but now that I have it it's become part of my daily life.
Tons of opportunity for innovation/products here. The only downside (besides the need for server voice processing) is that you are building a company on a platform. The whole Twitter and Facebook platform debacles showed the risks of trusting your business life support to a bigger indifferent company - who may just steal your idea and implement it.
--
Sent from my someDevice.
The way the ad depicts BK and Google "working together", to speak a favourable description of a BK product, suggests some sort of Google endorsement for BK.
Without that endorsement being approved, it's right that Google would act to stop the ad from suggesting or using an endorsement.
From my point of view, this is pretty risky for BK to do so. And either they didn't understand the risks, or they knew the repercussion would not be expensive for them, or they just decided to take the risk.
To give BK the benefit of the doubt from a strategic perspective, maybe the thinking was, "We know Google will react negatively or shut it down, but that reaction will also get coverage, and will promote talk of our ad even further. Thanks Google!" Still, I'm not marketing person, but that type of thinking, seems to me sort of like "poke the bear" for benefit, and seems quite bizarre to me. I feel it's strange if that is really how marketing people calculate.
Considering many dictionaries list it as "to google: to search in the internet with a search engine, e.g. Google® or Bing®", one could argue it has become generic, and the invocation isn’t actually "O.K. Google: ", but "OK, google: ", using it as imperative form of a verb.
That said, this argument is today very silly, as it relies on the letter-of-the-law interpretation, and wouldn’t fly in most courts yet. In a few years, with "to google" being used more commonly, it might actually.
It is an utilitarian mark. You can't trademark those. Or better, you can, you just don't get protections against utilitarian uses, like other manufacturers using the same mark for compatibility or companies using it to access the service.
Couldn't the same thing be done with Google and Alexa?
During installation, one could add his own voice, and then your command "Google, add this voice" would allow you to add another member of the household.
No, as I can use my fiance's iPhone and our voices are at least a fifth apart in tone, he has a northern accent, I have southern drawl.
E.g. if I say "hey Siri" while my phone has its screen off, it wakes up and waits for a query. If my girlfriend (or anyone else) says a query at this point it will respond to it. But if the screen is off and my girlfriend says "hey Siri", nothing happens.
That would still solve the problem the majority of times.
I don't even need to speak for that to happen. I just bend over with the phone in my back pocket and it wakes up, beeps, waits for me to say something (I say nothing,) boops, then goes "Sorry, I didn't get that."
What's evil, I believe, is that ultrasonic cross-device tracking technology. This was just cute.
But apparently what we don't know we don't care about when you tell us and we know but we can bury our heads in the sand.
汉保王 is, by contrast, wonderful
Every now and then someones device activates while I'm giving the example and all the talk of, "But its so convenient!" dies pretty fast.
Well then don't buy a fucking microphone that runs 24/7 in your living room, you dolt! Play stupid games, win stupid prizes.
Don't visit any of your friends who have bought it.
Don't go out in public.
Don't be a part of society.
Don't even SMS or email anyone since they might have one too.
The fight to reduce surveillance must happen even when the people who are fighting for privacy don't purchase the products that everyone else buys. If 99% of people have bought Google Home in the future, what good does it do to tell someone "hurr durr don't buy it?" No good at all.
The one thing we can do is create something like UL/Consumer reports for privacy. It is desperately needed. People automatically check this site to see what the exposure is of having a certain device in your home. I hope this comes to pass.
A trojan horse like Alexa and Google's device sneakily reverses this contract to the benefit of the corporation/government. People still feel the listening devices are "on demand" but in fact are being listened to all the time.
Education is really the only way I can see that any of us at the ground level can effect any change, here.
This is a clever trick. Get over it.
Second aspect, they have no voice recognition tied in to these assistants and that would curtail all this. Though not ideal it would be a high improvement from what we have currently.
But until we have customisable names for our assistants, voice recognition, then this and much more will carry on happening.
Do you have a landline and screen calls, well somebody could ring you up and speak to your assistant currently, that is an issue still.
For me I don't see things getting any better until we have the above tied into some additional user biometrics, be that array microphone tracking that is checked with visual recognition of some form.
Though a cheap simple solution would be a watch with one button that you pressed and spoke into and that was linked to the assistant. That would not be hard and work much better.