I can't be the only person to have identified the actual motives behind providing these assistants; that is collecting data for targeted advertising and profile building.
I'm still waiting for the world to realize that every Google product has the same motive --- data collection and privacy invasion. Lately, their focus has turned toward using security as a privacy invasion tool.
And not just Google, or per the article Amazon and Apple. While I'm long out of the Windows ecosystem (my Year of the Linux Desktop was a decade and a a half ago), it's been noted it has partly changed from a product you use to get your stuff done to a platform to sell you more stuff from Microsoft, plus whatever they might be doing or will do with its build in advertising.
And not just Google, or per the article Amazon and Apple.
Of these, Google is by far the worst in my opinion. Mainly because they don't sell any other product. *You* are their primary product.
Amazon is eating an increasing chunk of Google's lunch. They promote advertising in about the only way it really makes sense to me --- on their web site --- where I go to express an interest in buying a specific product.
partly changed from a product you use to get your stuff done to a platform to sell you more stuff from Microsoft
Yes, a predictable response to competition in the form it has taken.
The stats I've seen say they sold roughly 80 million of these this year. To be sustainable, this $15 device would have to cost in the neighborhood of $115 each.
If they can't sell them for at least a break even price, they are likely to do what Microsoft did and shut the whole thing down --- and leave you with a cheap brick.
Their initial plan and motivation was flawed. They were expecting consumers to use these things to just blindly order stuff from Amazon --- at a price most advantageous to Amazon of course.
Consumers may be pretty stupid but it turns out they aren't *that* stupid --- enough to blindly make purchases just for the thrill of doing it with their voice. Very few are using them for this.
While I agree that the Alexa devices would cost more if they were a normal product, $115 is too high. The Apple HomePod Mini sells for $99 and Apple is known to maintain a higher margin (average 35%) on their products than most other manufacturers. Assuming a similar cost for the Alexa (likely cheaper as the HomePod is a better speaker) and a lower margin puts the most likely retail price for an Alexa at around $70-75.
My $115 was a break even priced based on their units sold and reported losses in their Alexa unit for this year.
Adjustments can undoubtedly be made to bring this more inline with what the market will bear but it will almost certainly have to be much higher than $15 to break even.
Why would consumers care if they are sustainable?
If I get my value out of the product before it is bricked, I don't care if the company lost billions or even trillions of dollars?
I'm not sure I agree that losses on Alexa translate to higher prices, opposed to just cutting into profits. Most profitable businesses set prices too optimize unit profit times sales. That is to say, they they won't increase prices if sales will go down as a result. They aren't running a non-profit
The point I'm trying to make is that the cost don't typically roll down to to the consumer in the way that you describe specifically for the reason you call out. If Amazon is already doing everything it can to maximize profits and other sectors, they can't just make more money there to make up for a loss on Alexa. If there are leaving money on the table when it comes to e-commerce, they should raise prices whether or not Alexa is at a loss.
This is a separate question on whether you can get $15 of value out of a speaker before they cancel the program.
Uber might not have a long-term viable business model, but that won't stop me from taking a ride if I like the price. I'm a consumer not a investor.
If your point is that consumers might care about the long-term support for Alexa, then I agree that those consumers should be wary.
However, they don't need to worry about Amazon raising the price of toothpicks because alexa does poorly
You’re presuming too much competence on the part of Amazon. They’re not subsidizing those devices to fulfill a long term dastardly scheme where they make their money back and more.
They just screwed up lol and customers who took advantage won out.
Voice assistants are still great for kids who are too young to read or google things. A child can ask google assistant how to prevent spaghetti sticking together and get a better answer than a lot of parents would give them. They are really useful as media players too, especially the ones that can also control a tv and soundbar.
I have avoided voice assistant stuff almost entirely. I can't fathom what I would want it for. Then I house-sat for someone with a fully-connected google home. Turning on/off/dimming lights by voice and playing stuff on Spotify by voice is pretty excellent. Not excellent enough to make me forget the real goals of the device, but excellent indeed.
1. Music
2. Family Room TV control (we lose the remote a lot)
3. Interaction with the Home Automation system to a small
degree (mostly triggering scenes without needing to hit a button).
4. Weather
5. Timers/Reminders
6. House-wide announcements
On the few occasions we've asked questions, the garbage responses have trained us to _not_ do that.
Edit: Forgot to include my point. The only things on that list we actively do that could justifiably say they need remote access to work are music streaming and weather. And we could be local on music if we limited ourselves to owned libraries instead of streaming services.
I also use Siri on my apple watch for more personal things: todos, reminders, other me-specific interactions.
One issue I find to be huge but haven't seen mentioned in this thread is household assistants require shared accounts. So I can tell my watch "Remind me in 30 mins to buy that present for my partner" but if I did it on my Alexa they'd also see it. And thus we only use Alexa for truly household-wide things.
IMO a killer app would be voiceprint detection and an ability to funnel the speaker to their own account for things.
Alexa supports multiple profiles and voice print detection for profile owners. It's actually quit annoying TBH. I set it up a long time ago but it never really had any issues until recently. My wife's account is the one linked to the HA instance and it always voice prints me to my profile which isn't set up for HA control I guess. I'm not sure if they changed something about how HA stuff works on the backend or what, but I'll resolve it at some point.
I didn't realize it did voice print. Either I missed that or they didn't have it way back when we first got ours.
But yes, something where you could trivially say "these commands are generic, these commands are not" and it works perfectly would be the thing that puts a household voice assistant over the top for me.
We're close to that. Music, weather, alarms, and timers. Maybe an odd imdb or wikipedia trivia question now and then. I also use it to play as many pranks on the family as I can.
(btw, Alexa has become far worse at playing music in the last month or two... probably as an Amazon cost-saving matter.)
But, purely for those functions, it paid for itself in convenience in the first year. I'm happy with the purchase.
yeah... zero touch kitchen timers, music playing, and other doodad control is desirable. If you ever develop arthritis or get old you will see what I mean.
I'm not sure why techbros thought it would be the next multi-zillion dollar idea, though. Even with the adware they arent exactly making money from it.
This article makes me really curious about how Mycroft.ai will fare. Its makers have a different aim than the other voice assistance (privacy friendly voice assistant). I ordered one when they started their crowd campaign years ago and they're finally shipping. Between then and now my interest in voice command has plunged dramatically (I don't know why), and I wonder if their other potential users feel the same way. The fact that they aren't mentioned yet in this thread is depressingly telling.
Does HN have any recommendations for offline voice assistants?
I just want a voice assistant which can do basic things for me locally without connecting to the internet for simple tasks such as adding things to my calendar or a note, playing a song, opening an app and alike.
I'm using Android, but I'm sure others using iOS would also appreciate some suggestions.
I have a friend who uses Siri to great effect. Sometimes I use voice typing.
I have decided a few times that Android's Google Assistant is mature enough for me to try it out. But for two fatal flaws: it can't do anything useful, and I don't know how to speak to it.
The Assistant can't arbitrarily control other apps on a program, for instance if I want to say "log into my bank and check my balance on checking" there's no interface for that. We're left with mundane crap like "set an alarm at 10am."
Furthermore, I don't know the vocabulary of my Assistant. To be understood properly on the CLI, I've learned how to speak to Infocom's z-Machine, the Bourne shell, Cisco routers, hell I'm a full-fledged Protocol Droid at this point. But I have no idea what to say after "OK, Google!"
Sounds like they’re trying too hard to push a narrative with that headline.
Voice assistants didn’t fail at all, they just have a limited use case. Amazons delusions people would order products using their Alexa never were going to work.
35 comments
[ 2.6 ms ] story [ 69.1 ms ] threadI'm still waiting for the world to realize that every Google product has the same motive --- data collection and privacy invasion. Lately, their focus has turned toward using security as a privacy invasion tool.
Of these, Google is by far the worst in my opinion. Mainly because they don't sell any other product. *You* are their primary product.
Amazon is eating an increasing chunk of Google's lunch. They promote advertising in about the only way it really makes sense to me --- on their web site --- where I go to express an interest in buying a specific product.
partly changed from a product you use to get your stuff done to a platform to sell you more stuff from Microsoft
Yes, a predictable response to competition in the form it has taken.
I’ve got 3 Echo Dots and use them everyday. They do have limited use but people still buy them. For $15 USD, it’s a no brainer purchase.
As language models improve the device will become more useful but there’s no point in waiting.
It turns out that Amazon is subsidizing these things and is on track to lose $10 billion on them this year alone.
The price of everything on Amazon is higher because of these cheap toys.
no point in waiting
No point in buying if they aren't sustainable.
They aren’t losing $10 billion on hardware on a $15 device.
The stats I've seen say they sold roughly 80 million of these this year. To be sustainable, this $15 device would have to cost in the neighborhood of $115 each.
If they can't sell them for at least a break even price, they are likely to do what Microsoft did and shut the whole thing down --- and leave you with a cheap brick.
Their initial plan and motivation was flawed. They were expecting consumers to use these things to just blindly order stuff from Amazon --- at a price most advantageous to Amazon of course.
Consumers may be pretty stupid but it turns out they aren't *that* stupid --- enough to blindly make purchases just for the thrill of doing it with their voice. Very few are using them for this.
Why don’t you give that a little more thought.
Is this enough thinking for you?
Adjustments can undoubtedly be made to bring this more inline with what the market will bear but it will almost certainly have to be much higher than $15 to break even.
Why would consumers care if they are sustainable? If I get my value out of the product before it is bricked, I don't care if the company lost billions or even trillions of dollars?
Most don't --- it's typically out of sight, out of mind.
Most consumers don't care that credit cards consume 2.5% of the purchase price. But they ultimately pay for it.
Likewise with Amazon, their projected $10 billion loss on Alexa toys this year alone affects the price of everything else you buy from them.
If I get my value out of the product before it is bricked
And if you don't? It takes time, effort and money to fully incorporate Alexa into your home. The speaker is only the first part of the cost.
Personally, I'm reluctant to do this if I know the product is a marketplace loser on the verge of being axed.
https://news.ycombinator.com/item?id=33757288
Precisely the point.
$10 billion is a decent hit to the bottom line. They can choose to try and absorb it, but it typically rolls down to the consumer somehow.
If not, I can see a dead simple way to add $10 billion to profit --- shut down Alexa.
Neither of these options inspires a lot of confidence in the future of the product.
This is a separate question on whether you can get $15 of value out of a speaker before they cancel the program.
Uber might not have a long-term viable business model, but that won't stop me from taking a ride if I like the price. I'm a consumer not a investor.
If your point is that consumers might care about the long-term support for Alexa, then I agree that those consumers should be wary.
However, they don't need to worry about Amazon raising the price of toothpicks because alexa does poorly
They just screwed up lol and customers who took advantage won out.
These devices just collect audio data and send it to Amazon for processing. Without the backend service, they are useless.
[1] https://lecturia.org/en/short-stories/ray-bradbury-there-wil...
Edit: Forgot to include my point. The only things on that list we actively do that could justifiably say they need remote access to work are music streaming and weather. And we could be local on music if we limited ourselves to owned libraries instead of streaming services.
I also use Siri on my apple watch for more personal things: todos, reminders, other me-specific interactions.
One issue I find to be huge but haven't seen mentioned in this thread is household assistants require shared accounts. So I can tell my watch "Remind me in 30 mins to buy that present for my partner" but if I did it on my Alexa they'd also see it. And thus we only use Alexa for truly household-wide things.
IMO a killer app would be voiceprint detection and an ability to funnel the speaker to their own account for things.
But yes, something where you could trivially say "these commands are generic, these commands are not" and it works perfectly would be the thing that puts a household voice assistant over the top for me.
(btw, Alexa has become far worse at playing music in the last month or two... probably as an Amazon cost-saving matter.)
But, purely for those functions, it paid for itself in convenience in the first year. I'm happy with the purchase.
"By the way, I can..."
You are a light switch
"Oh my god"
I'm not sure why techbros thought it would be the next multi-zillion dollar idea, though. Even with the adware they arent exactly making money from it.
I just want a voice assistant which can do basic things for me locally without connecting to the internet for simple tasks such as adding things to my calendar or a note, playing a song, opening an app and alike.
I'm using Android, but I'm sure others using iOS would also appreciate some suggestions.
I have decided a few times that Android's Google Assistant is mature enough for me to try it out. But for two fatal flaws: it can't do anything useful, and I don't know how to speak to it.
The Assistant can't arbitrarily control other apps on a program, for instance if I want to say "log into my bank and check my balance on checking" there's no interface for that. We're left with mundane crap like "set an alarm at 10am."
Furthermore, I don't know the vocabulary of my Assistant. To be understood properly on the CLI, I've learned how to speak to Infocom's z-Machine, the Bourne shell, Cisco routers, hell I'm a full-fledged Protocol Droid at this point. But I have no idea what to say after "OK, Google!"
Voice assistants didn’t fail at all, they just have a limited use case. Amazons delusions people would order products using their Alexa never were going to work.