People (rightfully) pointing out that Siri sucks compared to competitors should realize that Apple is making a very serious tradeoff in favor of privacy.
It's a huge handicap when you're only allowed to train models on device using just one user's data, or only allowed to use various complicated privacy preserving training methods (I'm not talking about taking a user's data, removing the obviously identifying fields and sending it to the server, which does not really preserve privacy except for simple statistics). There's no way Apple can train the same super complicated deep learning models that Google uses, for example, when Apple can't just gobble up all their users' data. And all of the "provably" private methods for training models incur significant accuracy penalties.
So it's a little rich that users on HN will complain about Siri on one hand, and on the other hand complain about losing privacy.
In the end I think this is a fatal mistake for Apple: people love to complain about privacy, but time and time again reveal through their purchasing choices that they don't actually care, especially when there's usability tradeoffs. Apple would be better off vacuuming up all their users' data if they don't want to be crushed by Google/Amazon.
Of the 3 companies, Apple is likely the most privacy centric. However, they have a fuck ton of private data. You raise an interesting point; it may explain some percentage of why Siri is worse. Although, it has been in use longer and is still pretty shit. Apple sends a lot of telemetry and search data.
Also, I find it baffling anyone would want to put an internet connected always on mic in their house no matter who makes it. Insane.
As someone who isn't a fan of apples closed garden, I agree with your comment.
While I would like to see a model where they all share data to train better assistants, I know that will never happen.
Ontop of all that my side project works on voice recognition and trying to figure out what users want, damn it's frustrating not being able to grok at times, but I haven't quite gotten to the point where I'll take all user input, and put it through ml. (Lack of uses haha, go figure).
Also anything that Siri can do, you allow Apple to do. Because ultimately it's Apple that is sending the response commands back to your computer or phone.
Hmm, I think this will be a tough market to compete in for Apple. I think it's generally accepted that Siri is the worst of the assistants out there right now and that underlying technology will drive this entire new product. I really think Apple is going to need to decide between privacy and AI research as they are struggling to keep up and having a product like this will only make that gap more pronounced.
Exactly. I use Siri on my iPhone, also tried Google assitant app recently, having an echo at home too. Siri is way worse than the other two in terms of speech recognition and features. I felt Apple is better off to completely buy Google as the back end and simply focus on the hardware, where is their strength.
I think this is less a Siri device and more a home-TV device.
My understanding is that this device includes a screen. Right now I could use a 17" 16:9 iPad with a stand and maybe an optional detachable keyboard, as I find I'm using my 12" iPad Pro more and more for TV/movie viewing. I could use something like that for a bedroom nightstand.
I don't think Siri will be the focus of this product. I hope not, at least, because it definitely isn't the top assistant. Apple's strength is it's device media capabilities, and they should exploit that instead.
That's interesting. I've always thought echo was much much worse at speech recognition. It also does ridiculous things, like my toddler asked it what bread is, and echo proceeded to tell us about Bread, the rock band... Granted, Siri tends to just open up a web search.
Although their (commendable, IMO) limits on data collection may be a handicap, I think the extremely narrow Siri API is a bigger problem. At first I thought Apple was behind Amazon in terms of opening up the possibilities of using voice commands for more than a few simple things. But so much time has gone by and Siri is still so limited that I'm beginning to wonder if they have any intention of catching up.
Privacy versus AI research is a false dichotomy. Apple has been very candid about how they are able to maintain privacy while using data responsibly.
It's in the interest of Google, Facebook and others to get us to believe we have to give up our privacy, that it's the cost of advancement. But it's baloney. It lets them be lazy and continue to use us (the users) as the product for their real customer; advertisers. But it's just not true.
Privacy in the sense that you can train better AI systems when you have more data. Apple has many self-imposed and restrictive policies around collecting data and what type of data they collect, so they inherently have a disadvantage when it comes to building better AI systems.
Most of your data (ex. search history) is not reliably stored on your phone, and even if it was it would be much better to train on the server side instead of making your phone do all of the work. Also, they probably train using data from everyone, not just one user.
The problem may not be the Siri software, it may be the iPhone hardware. I believe the Echo has a pretty sophisticated microphone setup that is light-years better than an iPhone's. Apple is generally pretty good at hardware, so they could catch up quickly there.
A fair point. I never use the dictation, but I often try to use Siri when the phone is out of reach. My Echo understands me when I speak at a conversational level from across the room with the faucet running. If the dictation is spot on and Siri is still bad when speaking directly into the phone, then yeah, that seems pretty damning.
NLP quality is dependent on ASR quality (automated speech recognition, transforming audio into words) because you can't generally get meaning without words.
ASR quality is dependent on audio quality (plus, of course, a bunch of big-data-driven training) because it's harder to get the words right if you can't hear the sound.
Audio quality is dependent on microphone(s) quality and signal processing (higher signal-to-noise ratio).
And getting that to happen at far-field distances of 10-20 feet in noisy environments is not easy.
So... bottom line: the better the microphones, the higher the ceiling is on NLP performance.
Source: worked on Echo / Alexa for almost 3 years.
Does it really make sense splitting ASR and NLP up into distinct concepts? While I understand in the past they have solved different problems, it's my understanding that to get accurate ASR, grammar files or corpus can be provided to improve the recognition. I've also seen things like Siri autocorrect what is being transcribed as the sequence of words evolves.
Another example is how as an Alexa skill maker, you have to provide utterance / intent mappings. Is that just used to accurately classify intent / entities? Or is it also used to identify which skill to pass a user utterance to as a part of the alexa skill service because there could be very little variation between skill names or inquiries amazon is supposed to actually fulfill when a person is talking ??
> Does it really make sense splitting ASR and NLP up into distinct concepts? While I understand in the past they have solved different problems, it's my understanding that to get accurate ASR, grammar files or corpus can be provided to improve the recognition.
Yes, you are right, in that there are ways to blend the two (use data from one to improve the other). However, in the end of the day, the better the system can determine which words were spoken (using whatever technology), the better it can determine the meaning and the intent, and then decide what to do about it...
> Another example is how as an Alexa skill maker, you have to provide utterance / intent mappings. Is that just used to accurately classify intent / entities? Or is it also used to identify which skill to pass a user utterance to as a part of the alexa skill service
It is primarily used for the former (classify utterance / intent mappings for bootstrapping). Over time, with ML, the goal would be to help understand which "skills" apply to which intents. Unfortunately, today, that's not the case (in Alexa) and that's why the skill-specific keywords or names are still needed. From the user's point of view, it would be preferable to be able to say "Alexa, I need a ride to the airport" and have Alexa figure out whether Uber, Lyft, or the light rail service with a station two blocks from your house is the "best" option for you right now, based on price, availability, and your explicit and implicit preferences. Of course, the system would also have to allow you to specifically request a Lyft, if that's what you want.
(And, of course, it should ultimately be proactive and just offer to get you to the airport when it sees a flight in your calendar, of from having scanned the flight purchase confirmation in your email...)
Seems like it would be. I never use Siri speaking directly into the phone, because if I have the phone in hand, I just do whatever I need to do and skip Siri. The only time I try it is when the phone is out of reach, or in the car.
All these companies keep pushing voice input as the way to interact with their personal assistants, I personally feel that it would be 100x better to let me interact with them however I choose. In my experience, voice recognition is just not there yet, especially noisy environments, even if some companies are better at it than others.
If I could just type something into Siri I would use it countless times more. Most of my frustration comes from it misunderstanding me, often resulting in me pulling out my phone and typing in my question into search engine anyway.
Additionally, I do not want my business spoken aloud.
I never used voice input until I got my nexus 6p. Now it irritates me to no end when my wife wants to set a 15 minute timer for a power nap, and she has to poke around for 30 seconds, while I just turn my head towards my phone and say "ok google wake me in 15 minutes".
Or in the car, I'll pick up the phone and she'll prepare to say "don't start tapping on that thing while you drive!", but I just say "ok google navigate to [restaurant we're going to]".
Voice recognition isn't perfect, but a couple of years ago, android crossed the threshold where it's good enough for me.
I do similar things with my iPhone 6S. I say "Wake me up at 7" just about every week night, and then "Turn off my alarm" when I get out of bed in the morning. Realizing I could do this is why I started using my phone as my alarm; I wasn't about to fuss with the app every night and morning.
What baffles me is that opening some apps is not allowed without unlocking the phone. I should be able to say "Hey Siri, open NPR One" while washing the dishes. And I can, but it tells me I need to unlock my phone - but I used my voice because I have soap all over my hands, so no thanks.
Question, is your voice input experience mainly with iPhone and Siri? I find that voice input is fairly excellent for my Nexus 5X (less so for my Galaxy Tab).
Google seems to be doing a melded experience based on their Google I/O presentations. A lot of the third party experience will involve visual responses.
> All these companies keep pushing voice input as the way to interact with their personal assistants, I personally feel that it would be 100x better to let me interact with them however I choose.
Google Assistant is accessible by typing; it's built into Allo, and either built into Google Search on Android or the latter hooks into Allo if it is installed, I forget which.
So, I mean, you're still out of luck if you prefer interpretive dance as the interaction mode, but you aren't limited to speech.
If I use the Assistant integrated in Allo on Android and ask it to send a text to someone, I get a card with the contact information that asks me for the text for the message this send. (Same thing if I do it from the Google Search bar in Android.)
You seem to imply this doesn't work, but I'm not seeing it.
On my Nexus 5X, I can choose to type into the google search widget. Entering "when I get home remind me to take the trash to the street" will create a reminder, same as if I had spoken it. It's even pretty forgiving about typos.
Worse, using voice recognition allows companies to store your voice samples, that allows them to generate your voice fingerprints, that allows your presence to be detected once you happen to say something out loud in proximity of such devices.
From Apple's website: "processing is done by remote servers, it requires a data connection"
From Apple employee: "only collects the Siri voice clips in order to improve Siri itself"
Which is basically what Google says.
EDIT: both companies have promised "off-line" processing, whatever that means, but I don't know if any already does that. Note that off-line processing does NOT mean they won't store your voice clips on their servers.
It really does not matter if a single device does not work this way if the global trend is to push boundaries of privacy to closer and closer to non-existence.
I am afraid that soon it is hard for general public to avoid the situation where when they flush the toiled some unexpected device will recognise it and report to the headquarters.
No Spotify, no buy. Being locked into Apple Music is a non-starter for me. It's still mindboggling to me that Siri can't/won't interface with my Spotify account. My lowly Amazon Echo supports this just fine. In fact, I prefer it to the spotify UX/UI.
This is same reason why I refuse to use Fire TV, Android TV or Apple TV. I'll stick with Roku until the walled gardens fall.
I would suggest Roku make a digital assistant as well but it would probably come in 4 different models ranging from wholly unusable to outrageously expensive, have a purple bow tie, and need to be replaced annually with the latest model.
Roku itself is a walled garden. You cannot develop apps for them unless you are blessed by them. Last I checked the SDK was closed (as in, closed to even try out developing for Roku).
The BrightScript API for Roku app development is freely available, and any Roku set-top box or Roku TV can be put in developer mode where you can sideload your own channel. You can also upload private channels where users can install your content via a weblink -- they have to go through certification to be in the on-device Roku Channel Store. All developer docs and samplers are all on GitHub, see https://github.com/rokudev
Amazon has also shown that as a company, they're more than willing to engage in anticompetitive behavior when they feel threatened (for example, refusing to sell Chromecasts and Apple TVs, refusing to port Amazon Prime Video to competitors' platforms, etc.) I wouldn't put it past them to drop support for other services when it suits them.
Amazon refuses to even sell Chromecasts or Apple TVs and lies about it if you contact customer support. I've been told that "it's just out of stock", that "we had too many complaints", and even "Apple refuses to let us sell those products".
It's why I canceled Amazon Prime (Funny, after being a customer for over a decade, with Prime since it was offered, cancelling was a single click, with zero follow-up or attempt at retention. Not even a "Oh, hey, why did you decide to cancel after so many years?")
Just canceled too and was surprised how friction free it was. They sent just one follow up email so far trying to tell me about all the "benefits" I'll lose (which were things I never used anyways).
Right but there's no material difference between the two. You just have a choice between which master to serve. Amazon's platform isn't more open, they've just allowed other music services to operate on their platform.
Amazon approval is basically a rubber stamp. Can't say the same for Apple's. I'm also not aware of Amazon denying anything for competitive purposes. Amazon is happy to let Spotify on their platform, despite having a music service of their own, for instance.
I hope it is as impressive as the rumors claim as it'll help legitimize the "Speaker AI" market, but...what Apple needs to learn is that a product, no matter how advanced it may seem at launch, means nothing if they don't update and support it continuously.
The iPad and to some extent the Apple TV both have top of the line hardware. The software on the other hand...improvements every 2 or 3 years don't inspire confidence.
Updates to Siri once a year would kill this product.
Normally I would count this as an auto-win for Apple but the interesting thing about these speakers isn't necessarily the hardware itself but the many services that back them. I play with both Home and Echo a lot and I'm most impressed by how quickly Google is adding lots of useful functionality. Their documentation and evangelism for developing on their platform seems to be pretty good too. I think Apple will catch up quickly as they did with the Watch.
As far as I'm concerned, Siri is on death watch until Apple figures out a strategy to open it up. Outside of setting timers, creating reminders, setting alarms, doing a little dictating or mishearing directions, the product is stagnant.
I just don't see a lot of utility going forward unless they liberalize the service and allow other people to develop new experiences for users. That should also give them new data to mine and allow good voice UI patterns to bubble up.
Not sure how they should go about designing a framework for developers to plug-in to, though. It's a tricky problem. But presumably they've learned some things after 'sharing' some of Siri with Uber and Facebook and that could help them move forward.
I think that Apple has bet (correctly) that voice is a feature not a platform and that great hardware will make people buy into its services. Voice is great for a few things but very lousy for most things. Obviously it's all conjecture at this point, but my guess is that this product will augment the use of its devices for the things that voice is superior to GUI for (playing music and basic queries). Since Apple typically blows the competition out of the water with design and marketing, I imagine that this will do pretty well in this space (just imagine walking into an Apple store and seeing an IKEA-like display with a sleek home theatre and its speaker system on display).
I do think that the walled-garden is concerning since you can't use it with all music services, for example, but on the other hand, if its a premium speaker system that might be enough to push me the other way and switch to Apple music. The question is is the product compelling enough to get people to buy into the platform.
I agree that voice is lousy for many things, maybe even most things. And I think Apple is uniquely positioned to offer an end-to-end experience across devices (desktop, laptop, phone, watch, house/speaker) and that their design sensibilities could lead to a top notch speaker. I think the product will find buyers. But it's ultimate utility will remain in question until Siri improves.
I think part of the problem is that the things 'voice' is good for tend to have humans on the receiving end that then carry out complex, human actions. "Stu, could you schedule a dinner for me at a good local Italian place at 7:30 and invite Scott and Robin?"
Well shoot, that is a lot to unpack! It's a hell of a lot for a computer to unpack. It's a lot for anyone to unpack unless they know you and your circle of contacts. Carrying it out would require interaction with at least 3 people over a variety of mediums.
I suspect that by opening Siri up (And other digital 'assitants') it might promote the growth of infrastructure and services that would begin to make some of these more advanced queries a little more tractable.
I think that's very much the problem. Current voice systems are really just API-command-by-voice, maybe with a tiny amount of context awareness.
To be really compelling, voice needs to offer the whole AI concierge experience. At a basic level, it needs to be able to deal with queries like "Find me a good place to stay in Barcelona" and "Find me a new jacket". It needs to ask questions as needed, to operate with the initiative to search tens or hundreds of sites, and to recognise the useful data in the results.
With Google, Amazon, and the rest, search is becoming more and more of a problem, not less. For non-trivial searches, finding good products and/or reliable information can be incredibly time-consuming.
So currently voice is a bandaid on top of search tools that aren't progressing much, and may even be regressing. Voice has to solve the search problem first before the recognition and context awareness problems really become important.
It looks as if the industry is to trying to do this the other way around. I'm not convinced that's going to work. It works up to a point, but the point isn't as advanced as users expect it to be, and the overall experience can be disappointing.
Apple's prowess is in designing physically attractive objects, and easy to use UI. Voice has very little UI, and it has even less use for physical appearance. That's not to say that how these things look isn't important. It is. But it's nowhere near as important as it was with a phone.
My argument is that voice isn't all that useful in general and that if Apple creates a much higher quality product than Echo, for example, that people will want it because it's a very nice speaker that looks really cool and can play music, give the weather, and set a timer by voice, which covers the majority of use cases.
In other words, a voice platform that works 10x better than another may only be slightly more useful to the average consumer while if the actual speaker is 10x better, it might be 10x more compelling.
Apple already opened up Siri. Third-party app developers can now add their own Siri integration (though I don't know how extensive this is, having never looked into the details myself). But it seems like not very many third-party developers are actually bothering to do so.
There's just a few things you can do. Call someone, send a message, hail a cab, ask for directions, stuff like that. Six scenarios in total. You could try to WhatsApp someone with Siri, probably that'll work. Though I hate using voice commands I never tried it.
I'd really love if Apple integrated the Workflow app they acquired into Siri. I think it could be a straightforward way to let Siri do more useful things than just setting reminders and asking about the weather. They also have improve the voice recognition big time since Siri has real difficulties understanding if you have an accent.
I still don't understand the desire to have any sort of electronics constantly listening for commands. In the balance between convenience and privacy, to me, devices like this don't seem to add enough value to draw me in.
Speaking personally, I've never managed to work out any real, practical concerns with devices, services, and situations like this. To start, I haven't seen any reason to be more concerned about Google Home than I am about my Pixel phone, which is not only "constantly listening", but also accompanies me everywhere I go outside my home.
I never leave my phone somewhere else when I'm at home, so... what's the difference if I add an extra microphone/speaker to the system already in-place?
Even besides that, I've never managed to work out what people who cite privacy concerns are actually worried about; I've never managed to come up with any scenario that both seems worrying and seems to me like it has any plausible chance of actually happening.
If we imagine a hypothetical counterfactual world where Google records everything ever spoken in my home or near my phone, retains these records forever, and analyzes them and uses the analysis to choose which ads to show me, and promptly responds to any requests by law enforcement for recordings with no hesitation or review process, and recordings are available for people at the company to listen to... what would happen that I would be upset about? Fewer poorly-targeted ads? Law enforcement having a slightly more-trivial time than usual locking me in a cage if I happen to offend a law-enforcement officer? Someone I'll never meet listens to me having sex?
It might be the case that I just happen to have a life situation that's uniquely stable and not susceptible to being fired for political opinions or whatever the risk is... but I don't think my attitudes and risk evaluation here are really all that unusual. I speculate without evidence that most people just really don't care if they're being watched unless it's going to have practical consequences.
I recognize that Google isn't recording everything said, etc. However, with video (tv, internet) ads prompting these devices to search [1], its not something I have an interest in participating in.
Ultimately, for me, it boils down to a lack of benefit for me, and a lot of benefit (current and potential) for Google and these other ad-drivers to track my decisions more than I really want or need them to. As it stands, I live a fairly ad-free life, either through payment or blockers --- so the targeted ads are less of a concern. Law enforcement concerns are also extremely limited, as are political (null).
I just don't see the benefits. If I need to know the current weather, I'll look outside. If I need to search something, it'll never be so important that it can't wait until I sit down at my computer.
I guess, overall, I'm less concerned about the privacy aspect as I am the actual personal benefit of these devices. In my mind I relate it to the Dash buttons.
Following house of cards? US elections? Remember WW II? Where they had just a tiny bit of info on a specific group of people (their faith, and registered address) and used that to the max to commit genocide on them?
Think that'll never happen?
Question to the Muslims on HN: how do you feel about having all your personal data stored by corporations and the government?
The device is maintaining a short (<1 sec) buffer at all times. The device is Echo; Alexa is the cloud-side service, which is definitely not getting the audio until the wake-word is spotted and then streaming begins (backing up to the beginning of the buffer).
Incorrect. Police asked Amazon for a murder suspect's Alexa history, and Amazon initially refused until the suspect/customer gave permission. It is yet to be proven if anything was recorded without the wakeword.
All we know is that the recordings were requested by the prosecution; we have no idea what, if anything, is included in them. In particular, we do NOT know if anything other than the wake word, followed by a few seconds of query, is included. Everything else we've learned, including Amazon's own statements on what is recorded and transmitted, suggests that nothing else is included.
Are these home speakers the new smartwatch? A product that every company has to make because it's "the big new thing" but the actual functionality is questionable?
I don't understand why I'd want a tiny speaker sitting in my house listening for commands, when I already have that on my phone and it can provide a display of the results, along with voice.
There was a great "Ask HN" thread a while ago on this and I read everything thinking "oh good, power users will tell me what I'm missing here" and when I was done I was left thinking "I am not missing anything here". In their current form, they just didn't seem to do all that much that would make any measurable difference in my life. I have an echo, it was plugged in a for a week and everyone forgot about it.
I have a Google Home, rather than an Echo, but the biggest changes are the fact that I listen to a lot more music and I have it hooked up to a Logitech Harmony Hub so that I rarely need to hunt down the TV remote. "Hey Google, Turn on the TV // Pause Roku // Volume down" are pretty useful shortcuts I have configured.
I was skeptical but bought a cheap one because I thought it would be fun to hack on. Turns out it was more useful than I imagined. We listen to more music in our living room now that it's become frictionless to play something (I'm not an audiophile but I can't listen to music on cell phone speakers). When I tell the kids it's bedtime in ten minutes and set an echo timer, they actually go upstairs when the timer goes off (probably some weird psychology going on there, but the living room telling them it's bedtime carries more authority than dads phone). The killer app for my wife and I has been OneBusAway. Being told when the next bus is coming while getting ready in the morning is awesome, much better that fumbling around with their app.
Obviously your use cases may vary, but with my echo I was pleasantly surprised at how useful it was. I was given a smartwatch for free and was nevertheless disappointed at how useless it was. I can pull my phone out and do stuff with one hand, whereas on a watch you need to hold one hand in place while the other hand operates the device, meaning it's harder to use and less convenient.
Going to be hard to outdo Google and Amazon with the dogshit that is Siri. Ok Google and Alexa are lightyears ahead of Siri in terms of voice recognition and response.
Macrumors says they are going to differentiate themselves by deeply integrating with the Apple ecosystem and having a better speaker.
These aren't real selling points.
Google Home integrates very well with Google products (out of the box I thought to try asking it to 'play the latest video from [youtube channel] on my shield', and it worked, exactly as expected.
As far as speakers, Amazon started with a device that had very good speakers, and has since downgraded to a puck.
FWIW, the speaker on the original Amazon Echo isn't very good.
(I also own small Bluetooth speakers from Sony and Anker, and both are appreciably better for music — enough so that I go out of my way to use them while cleaning or whatnot, even though the Echo is right there.)
I keep wondering why they don't open Siri up to the command line or applescripts. It seems like that would be a great way to grow the Siri ecosystem. Does HN have any ideas why it is so locked down?
It's not the hardware - it's the software. I went all in on Amazon Echo and Dots. I have them in my living room, office, bathroom, bedroom.
The ecosystem that Amazon provides is far richer than Apple can hope to provide. Google does seem to be better positioned.
Even if the hardware is amazing, I don't want Apple music (I have prime music) the voice assistant is no better and I have serious doubts about the integrations that will be available.
Whats going to make this DOA is integrations, it's not in Apple's DNA and its going to be the best sounding, most beautiful, most seemless voice assistant nobody wants. I hope I'm proven wrong.
More evidence today's Apple lacks vision. These products are niche. Some people love them I know. But look at the sales. They are not appealing to the mainstream. The future of digital assistants is not dedicated devices in every room of your home. I'm not sure what it is but it's not that.
I believe the best way to interact with Siri is through the Apple Watch, something that is always attached to you and doesn't require you to shout across the room. The Watch also connects to Wifi, and can work independently of the iPhone, allowing Siri, iMessage, calling etc. Also, the siri speaker has to deal with the possibility of multiple users using "Hey Siri" while the apple watch can be tied to an individual person.
I've had a hard time finding much useful with these devices - anecdotally even my friends have been ditching their Echo's.
I've ended up dropping my usage to setting timers and turning on/off lights, both of which don't work with Siri in MacOS.
I'd hope that they'd get their own ecosystem in order before adding yet another gadget, but that's not happened time and time again under Cook's leadership.
So I really have tried to use Siri. Really. I want it to work. My home ecosystem is all Apple (for the user facing devices). However most of the time with anything beyond "set a reminder or set an alarm" I give up and just do it the old fashioned way. Part of the issue is that the backend search results are just no where as good as Google. IIRC Apple is using Bing for that..Ugh.
Google's Assistant isn't any better at this sort of inference. I have my wife's work and cell number and I get asked any time I use it which one though I'm reasonably sure I have never once caller her work number.
The Google Assistant happily asks me if I want to finish researching "Care Bears" because I once searched for them on Amazon using Firefox in Incognito Mode on a friend's computer but it can't seem to figure out that I always only ever call my wife on her cellphone.
Uhuh... I tried to ask it for directions to Mom's tonight in the car and then for my mom's address by her full name and it failed on both attempts. It did however manage to find her Pinterest account.
Siri understands contact relationships if you fill out the "related names" in your contact cards.
Go to your own contact card, and then set a related name for mother to your Mom’s contact card. The next time you ask Siri to “Call Mom” it will do it.
It’s unnecessarily complicated, it should be able to infer which one, but there’s the workaround.
There really isn't one. But normal people (i.e. non techies) use these things for small "factoid" type situations. Weather, sports scores, stuff like that.
It takes longer to ask a thing what the weather is than looking at my phone, but that's just how it is.
Siri is worst for indian English Accent like me, Could not even compare it with Google. Google is light years ahead , it can recognize my english 100% and most of the times even transliterated tamil spoken in English (another local indian language)
Apple has always refined things that are almost working but clunky and awkward to use. Mp3 players were terrible for non-nerds for example. Browsing the web on a palm pilot sucked. Apple's problem now is that their competitors are building pretty good stuff. Echo is a great combination of hardware and software. Surface laptops and tablets are quality.
VR/AR is an area where everything is clunky and awkward to use and Apple could probably build something really amazing here but the market size is tiny and the use cases are outside of apples wheelhouse (industrial applications and gaming right now).
Oh great... another self-driving car project... /s
To me, this just feels like another jump on the bandwagon. I have an Echo and, while it was awesome at first, with all the skills and everything it feels less amazing than it did when I first got it. Now, I feel like it mishears us all the time, it doesn't give the answers we'd normally get unless we say things in a very specific way, and all the cool tricks ("Who is the mother of dragons?") are just gimmicks now.
Unless Apple can actually make their version useful and able to understand questions that aren't formed in "robot" as the language, I don't see the point of this.
I've found the Home useful. All I can use the Echo for is to set multiple simultaneous kitchen timers and to play music while cooking. The Echo just doesn't seem to be able to handle natural language well or there isn't support for what I'm asking.
Multiple kitchen timers is by far the most common use case I have for my Home. Not coincidentally, it's also my #1 complaint. Google, when a timer goes off, please please just tell me the name of the timer that is going off. Does the Echo do that?
Unlike the self-driving car project, a voice-driven assistant is not an option for Apple. It's an essential piece of the next-gen OS, and if they get locked out of it, the iPhone is not going to be enough to save them. They need to be in the car and on the TV (and everywhere else), and voice is key to that kind of ubiquity.
They are convenient voice command lines. It's faster to speak your timer, ask the weather, play music, or remotely turn off lights than to fiddle with your phone.
Sure... if you have them memorized. Most of the time, the Echo doesn't understand what I'm saying unless I phrase it exactly the same way each time and it pretty much is me learning through trial and error what lines work and what doesn't. Siri, at least on my phone, seems to get natural language better.
Apple was thinking about doing driverless cars of some sort after Tesla and Google. They seemed to be taking employees from Tesla and Ford for a while, but recent rumors are that the project has been halted.
They also registered the domain apple.car
The grandparent is likening Apple jumping on the bandwagon after competitors' cars to them jumping on the bandwagon with the speaker.
It feels like this is jumping on a bandwagon. How many competing smart home speakers do we really need? It's just like the situation with self-driving cars. How many players are there now? Waymo, Intel, Uber, Apple, Yandex, Lyft, Tesla... how many of these do we need?
I don't understand this mentality. Apple often enters into "crowded" markets and then establishes either a clear winner, or a strong contender. The market's existence removes any question of whether people are buying, so... what's the problem? They just shouldn't do it because other people make them already?
Lots of others make laptops, smartphones, headphones, set-top boxes... and yet, and yet. They still turn quite a profit.
We "need" one one supplier of self-driving tech for each large global car company. That's how previous technology transitions in automobiles have worked: usually everyone eventually figures out to build "it", whatever "it" is.
Unusual? _If_ the 'outdo' of this article really is OUTDO, it would be a repeat of what they did with MP3 players and mobile phones.
Thats a big if, though. Services aren't Apple's strongpoint, and their stance on privacy means they may have to do a lot of the processing in the device, and limit their ability to rapidly improve their offering.
You have a point. I was thinking of the 'speed' of the iCloud web interface and the 'speed' of the iOS update check (it currently takes about 4 seconds for my iPad to report that its software is up to date. When there are updates, that can easily become 10 seconds. I find that slow for what could be a single http request that returns one or two _bytes_)
I'm guessing the update checks are somewhat more involved than they appear. It may actually be validating that the local device is eligible to receive the update using some criteria, rather than simply saying "what's the latest version?"
What do you mean? I use iMessage probably hundreds of times a day (to my chagrin, but I digress). The Music Store has been overtaken by streaming services, but it made a titanic impact on the market -- and as much hate as Apple Music gets, it's actually a completely adequate service.
You're kidding, right? iMessage has massive scale. Several years ago, Apple revealed they were processing "several billion" iMessages per day, and it's hard to imagine that number hasn't simply grown since then.
As for the iTunes Music Store, I could believe that it's not used as often as previously because of the popularity of streaming services, but Apple also runs one of those streaming services (Apple Music), and they run iCloud Library which gives you access to all of your purchased music on all of your devices, and of course they also run the iOS App Store and you can't argue that doesn't have massive scale.
Agree, they are big but that's because they are pre-installed on hundreds of millions of hardware devices. Not because the services themselves are particularly good.
At least see the product first, before making such a statement. Apple does quite well by entering a market late (digital music, phones, etc.) and then dominating it by making superior products and user experiences.
If Apple releases the product and it's just a "me too device" with nothing new, then your statement is reasonable.
> Apple Inc. is already in your pocket, on your desk and underneath your television. Soon, a device embossed with “Designed by Apple in California” may be on your nightstand or kitchen counter as well.
It's a huge handicap when you're only allowed to train models on device using just one user's data, or only allowed to use various complicated privacy preserving training methods (I'm not talking about taking a user's data, removing the obviously identifying fields and sending it to the server, which does not really preserve privacy except for simple statistics). There's no way Apple can train the same super complicated deep learning models that Google uses, for example, when Apple can't just gobble up all their users' data. And all of the "provably" private methods for training models incur significant accuracy penalties.
So it's a little rich that users on HN will complain about Siri on one hand, and on the other hand complain about losing privacy.
In the end I think this is a fatal mistake for Apple: people love to complain about privacy, but time and time again reveal through their purchasing choices that they don't actually care, especially when there's usability tradeoffs. Apple would be better off vacuuming up all their users' data if they don't want to be crushed by Google/Amazon.