Hacker News new | past | comments | ask | show | jobs | submit login
After a decade as Apple’s assistant, Siri still hasn’t figured out the job (theverge.com)
146 points by 1cvmask on Oct 5, 2021 | hide | past | favorite | 206 comments



> is it a feature focused on search and retrieval, or an assistant that carries out complex tasks?

I'd settle for an assistant that carries out simple tasks.

I don't expect much from my robot assistant, but the number of times throughout the day that she gets it blatantly wrong has become a running joke in our household.

A few recent examples:

    Me, holding my Apple Watch near my face: "Hey, Siri, drive to Kroger <there's one I go to all the time that is 10 minutes from my house>"

    Watch: "Now playing <some band I've never heard of, that isn't in my library>"

    Simultaneously, iPhone: "Getting directions to Kroger <shows results that are ~700 miles away>"
Or:

    Me, to my iPhone: "Hey, Siri, start a yoga workout."

    iPhone: <starts playing a Yoga video on Fitness+, instead of starting the plain Yoga tracking workout I do every day>
Edit:

I just remembered another one. I was mowing the lawn on Sunday and wanted to start an outdoor walk. The mower was off, and I had my AirPods in:

    Me: "Siri, start an outdoor walk"

    iPhone: "Ok, I've cancelled your alarm for tomorrow"


I had this experience over the weekend:

Uses Siri on Apple Watch to start a timer Uses Siri on Apple Watch to start a timer "Hey Siri, set a timer for 3 minutes" "Sorry, you don't have the Timer app installed"


This gets to a problem I’ve had with Siri once it rolled out to multiple devices. Each device has a different set of capabilities. Sometimes this makes sense (eg I don’t expect my TV to be able to place a call. It doesn’t have a microphone, nor a camera.), but many times it’s just bizarre. For instance, I can’t set an alarm on my laptop. It has a clock, yet I can’t do clock things.

It seems like some basic feature (eg clock like features) should just be built in, and where that’s not feasible, perhaps some sort of device handoff, which essentially happens on the watch.


When I'm a cynical, I'm thinking bad search and bad implementations (to me) are really good implementations, the difference is that the goal is not my goal.

Examples: * when starting to type, Chrome address bar preferring to go to Google search, instead of the friggin site I've gone to a hundred times already! * Google maps, search seemingly doesn't deterministically remember my searches. If I start typing the same search term I did yesterday, well, odds are I'm looking it up again! Show that. * Windows start menu... oh don't get me started on that. :)


I'm having trouble getting Siri to even understand me. I have a HomeKit scene called "Apartment lights off" (it turns off all the lights in the apartment surprisingly enough) and I often ask Siri to apply that scene. The amount of times Siri completely misunderstands me for something strange baffles me.

"Upon lights off – Sorry, I don't understand."

"Obama lights off – Sorry, still don't get it."

Sometimes it just picks up the "lights off" part, prompting me "Which room?" And when I say "Apartment lights off" again, Siri will then reply with "Sorry, I can't find that scene in the current home." At that point, I need to close the Siri prompt and do it again. If Siri understand me that time, then I will actually get that scene applied.


I am suspecting the problems you have are due to the name of the scene. It appears certain keywords do overrule locally-defined names. Whenever I put an address book entry for a taxi company with the name "taxi" in it, Siri isnt able to call it because it assumes I want to search for a taxi nearby. A similar thing might be going on with "apartment" and homekit. Frankly, a bit cynical, but by now, I am suspecting some intern put a rule in which contains "apartment" and has a very high priority, or somesuch...

Not directly related to Siri, but VoiceOver icon labels are also getting worse and worse over the years. For example, the weather app. In the beginning, when you scrolled with the VoiceOver cursor through the hourly forecast, it would say "Sonnenaufgang" and "Sonnenuntergang" (Sunrise and Sunset) at the appropriate places. Starting with iOS 12 (or was it 13?) the spoken label changed to "Sonne aufwärtspfeil" (sun arrow up). So someone went in there, decided that the nice description from earlier wasn't good enough, and replaced it with a literal description of the icon, ignoring context altogether.


> “I am suspecting the problems you have are due to the name of the scene.”

yah, for siri, i’ve found simple nouns (apartment) better than phrases (apartment lights off) for naming things in the home app, and for “all lights”, you don’t even need to create a scene (as others have pointed out). for rooms, you can just say “turn kitchen on” and it will turn all lights in its kitchen room on (which may or may not map to all lights in your actual kitchen, depending on how you set it up).

oh, and siri will refuse to do something if you curse at it: “hey siri, turn on the lights, you asshole” ominously responds with “sorry, i can’t do that (…dave)”


Except I don’t have this issue with Google Assistant. It guesses the correct word/phrasing nearly every time. Siri, OTOH, fails much more than it succeeds.


"All lights to max"

"Which phone number for Max?"

I wish I could train myself to say a different phrase.


Try using groups instead of scenes. Also, key by room, or even use blanket universals (rather than say apartment).

“Turn on all living room lights” works. Also, “turn off all lights” will flip off everything in our house.

Groups are awesome. I have 3 blinds in the north of living room, so I called the, north west, north center, and north east. But they are in the north group, so I usually put them all up at once by saying “open north blinds.”


The problem with "turn off all lights" is that if I have some third-party devices in HomeKit that are not actually lights but are registered like lights, they go off as well.

Also, it's a studio apartment, so there's not really "rooms" strictly speaking.


You can always assign them to rooms even if you just have one room. But I think groups would work for your case. Like say put all your actual lights in the actual group. And then say “turn on actual lights.”

Scenes..I can’t really figure them out. They seem way too complex for what I want to do, which is to turn things off and on.


Grouping won't work at all since if I group the accessories, I will lose control of the individual lights.

Really, this is what scenes are designed for but Siri is just too stupid to understand what they are.


Scenes are activated, they don’t have direction, they don’t work at all for turning things in and off, instead you have to create separate on and off scenes.

HomeKit should really support overlapping groups, it is a feature I also miss. But as a hack you could use rooms as a secondary group mechanism since you don’t need them to represent actual rooms.


How has Doonesbury not put Siri out of her misery?

I started using this garbage with the 4S and while there have been under the hood improvements of onboard processing, etc, the core product has never gained the implicit context awareness promised and needed. As far as I can tell, it's a checkbox zombie "feature" that never met the usability threshold to become something for normal people but organizationally aged into slow development maturity.

https://www.computerhistory.org/revolution/mobile-computing/...


I’ll bite.

Me to watch: Hey Siri, call so-and-so on Bluetooth

Siri: calls so-and-so with Bluetooth selected but nonetheless the audio comes out the watch until I touch the phone.

Siri on Apple Watch is generally very weak.


Recently I went through the trouble to create lists to use Siri to add shopping items. I say, “add mustard oil”. Siri then adds “mustard” and “oil”. No problem, I’m smart. So I create a custom word in my phone and train it on my pronunciation. Try again, “mustard” and “oil”.

Guess I’m going to be shopping for “oil of mustard” then


I have also found Siri’s voice recognition for shopping list items lackluster, especially compared to Alexa. I’m guessing it has to do with Amazon having a better model of actual subjects/products due to having a storefront with said items.

It’s really bad when I’m at the store looking at something in my list scratching my head over what “Silicon Matt to sit rock song” is supposed to be without any way to hear what was actually said or give feedback that what Siri did was wrong.


"Hey Siri, is my phone still charging?" - "The current volume is 43%".


Siri can't start an outdoor walk, silly. She doesn't have legs.


Are you Scottish by any chance? https://www.youtube.com/watch?v=NMS2VnDveP8


I am not :/


Having multiple things actively listening for "Hey Siri" seems like a problem you've created, honestly.

I have my phone set to only engage Siri with the long button press. It saves a lot of trouble.


Blaming the user for the lack of simple coordination between devices when they buy into your ecosystem across multiple devices and attempt to use the advertised features doesn't seem like a great product strategy.


The user understands this limitation, and rather than lighting a candle is cursing the darkness.

It's redundant and counterproductive to have multiple devices listening for the code phrase.


Is it really blame if it's simply a technical deficiency?

I imagine that if I put 10 iPhones on a desk and say "Hey Siri", that wouldn't work out too well. Doesn't mean we have to start "blaming" people, companies and things.


The phone and watch are advertised as intergraded with each other so this does not happen. Just like answering FaceTime on one device is supposed to stop the others from ringing. It is a feature of the watch and iPhone to figure out which should be listening and it is broken.

To the article’s point, Amazon’s has no problem Determining which should answer a question if they are all the same account piled on your desk.


And yet there are alarms set on the watch which are separate from alarms set on the paired phone. It’s really madness.


We are not talking about ten phones on a desk. We are talking about a phone and a watch, which are explicitly designed to and advertised as working together. Surely you see the distinction between your thought experiment and the actual scenario in question?


Surely you understand that the scenario is merely an attempt at illustrating the difference between blame and technical deficiency.


the entire point of these assistant technologies is supposed to be that they seamlessly integrate with the environment. If you have to do long button presses or avoid this or that you can just open Spotify yourself and click play. Given that the benefit you get out of this stuff is likely marginal to begin with failing in 5% of cases is already way too much.

At this point to me it seems like most of it is basically toy technology without any real benefit unless you're visually impaired or unable to use your hands or something. People went from voice messaging to texting with other people because tactile inputs are often faster, so it's always puzzled me what the point is of talking to machines who can't even understand things properly half the time.


>the entire point of these assistant technologies is supposed to be that they seamlessly integrate with the environment

No. The "entire point" is to provide an easier way to do some interactions. On this point, they work super well. They're just not magic or perfect.

However, it definitely IS useful to state the point that way, because it makes it easier to argue that they're pointless toy technologies, thereby allowing you to feel superior.


Multiple devices are supposed to coordinate amongst themselves to decide who takes the request:

https://support.apple.com/en-us/HT208472


As a HomeKit user, I was looking forward to iOS 15, thinking there might be some incremental improvements to Siri’s handling of HomeKit commands. It wasn’t bad before, but it had room for improvement.

Unfortunately iOS 15 significantly deteriorated Siri’s understanding of HomeKit in some situations. Here’s a few reproducible examples:

‘Hey Siri, shades down’: sets the shades to 99%, i.e. a 1% adjustment. I now need to say ‘closed’ rather than ‘down’ to close them.

‘Hey Siri, close the living room shades’: responds with the current shade status. I need to switch the ordering to get things working (Hey Siri, living room shades: closed)

Did someone delete a mapping file of common HomeKit voice commands to actions or something? At this point, I’d almost prefer no AI, and just a static command list. I guess I could create that with shortcuts, but with 20 or so accessories that’s a big chore to set up.


This has been driving me insane since iOS 15.

Heavily invested in HomeKit (to the point of having several HomePods around the house), and it feels like something changed and much of it I can't quite put my finger on.

But a few that I have been able to identify:

- Something has changed about how it handles figuring out which device you are talking too. I have had a HomePod in a different room go off (screwing up the room awareness), or my iPhone going off even when I am right next to a HomePod.

- All of my HomePod pairs (having 2 linked together) switched to the opposite HomePod for which one speaks, and there is no clear way to change this.

- Every action has slowed down, I can't figure out why but most of the time I get "one moment" while it tries to turn in lights or whatever.

- They added the ability for the HomePod to turn on and off the Apple TV. Which is great in theory but I have my own custom TV configuration hooked up to Logitech harmony and it will constantly ask me "do you want to control x or y" (one being my custom and one being the Apple TV depending on the room I am in). I want to disable this but I can't.

Edit: All I really want is a log. Please. A log of automated events that happen (I still don't know why my lights will randomly turn off... something is triggering it and I can't figure out what) and a log of what it thinks it heard and tried to do so I can address the "nothing with that name or function is found" or whatever it says when I say "Turn off the fan" and it doesn't know what I meant, and then I try again and it works.


Glad to see I'm not the only one. I feel like a bunch of prior working commands have gone missing or are interpreted entirely wrong. Siri doesn't respond properly to "Turn the radiator in living room off" but when you say "turn heating in living room off" even responds with "turning radiator in living room off". As a consumer, this is downright frustrating.


Could this be why?

https://www.engadget.com/ios-15-siri-on-device-app-privacy-1...

"But with iOS 15, many common Siri requests will be processed on-device, so no audio will ever leave your phone."


Homekit sucks in general. Even without Siri the UI is unresponsive, broken, missing obvious features. The UI needs to show at least 3 states (maybe more) but only shows 2, on and off. Often I'll click a light and get ZERO response from the UI. IIUC it's waiting for the light to respond with its state but a UI that gives no immediate feedback is useless. It's infuriating that Apple who wrote the interface guidelines about instant feedback didn't apply them to their own software.


“Hey Siri, close the living room shades” works for me. The problem I have is that we have both sun shades and black out blinds and Siri treats the words “shades” and “blinds” as the same word, forcing me to create special group labels for each so it can distinguish them.

That said, it does feel like magic to be able to easily control all the lights and blinds with voice commands and Siri usually gets it right.


Coming from Android, I used to use the Google assistant in hands free situations. Like when cooking or driving. I've found that Siri is TERRIBLE in these situations, because its more likely to punt to give you some links to click on and say "I found this on the web".

I just asked Google and Siri "what temperature do I need to cook fish to". Google came right back with 145F. Siri put up some stuff she found on the web, and didn't give a verbal answer.

I like almost everything better about the iPhone, but I really wish I could replace Siri with Google Assistant.


The only time I found Siri useful is disabling the bluetooth in my MBA. I have a setting that whenever that BT mouse is connected, the trackpad is automatically disabled. Sometime I forgot to bring it with me when I brought the MBA outside or other room. I asked Siri to disable the bluetooth so I can have the trackpad functionality back.


145F is beyond well-done for fish

Edit: I must refrain myself from changing this comment into a pun about it seeming fishy.


related, I read recently that the temperatures were lowered for some meats for internal temperature from 165.

EDIT: ok, here's the chart:

https://www.fsis.usda.gov/food-safety/safe-food-handling-and...

some things are still 165, but some are now 145. ham in some cases is 140.


Came across this last year, which intuitively made sense to me--definitely have had chicken cooked at lower than 165F and didn't get sick.

I've not seen charts for other types of meat though.

https://www.seriouseats.com/the-food-lab-complete-guide-to-s...

"The upshot is: Food safety is a function of both temperature and time. What the USDA is looking for is a 7.0 log10 relative reduction in salmonella bacteria in chicken. That is, a reduction that ensures that out of every 10,000,000 bacteria living on that piece of chicken to start, only one will survive."


I think you realize this already but having getting salmonella from undercooked chicken is not a guaranteed thing.


but in the chart, poultry is still 165.


I'm convinced they've swapped out the microphones used by Siri for cheaper and cheaper ones over the years because it no longer seems to be able to pick up your requests across the room which it could at the launch of Hey Siri reliably.

Presumably because the whole AI assistant thing kinda turned out to be a fad and not as world changing as people were thinking around the time Siri/Alexa launched.

Turns out the thing it's good for it setting timers not ordering soup like Apple originally envisioned. https://www.youtube.com/watch?v=EP1YAatv1Mc


Edit: forgive my enthusiasm. Voice assistants seen as command-lines make me excited.

No it's not! The issue is that we as developers can't actually innovate on it. Look at what Apple did in its heyday with the App Store. Now it's commonplace, back then it was mindblowingly innovative unicorn goodness (I'm a bit biased :) ). And I understand, you can critique the App Store for many things, but the iPhone would be worse without it.

And they didn't create an "App Store" or any other free extensible model for Siri. They should have, because people are on aggregate more creative than any big company including Apple -- thanks to survivorship bias.

What Siri is actually quite good at is interpreting single words or small sentences. So it would be a perfect spoken command-line interface. I have had genuine situations where I needed to talk and couldn't type, but apps aren't allowed to make use of Siri, so I can't talk to apps, I need to type.

Siri and other voice assistants are amazing when you look at it as we look at a command-line.


> And they didn't create an "App Store" or any other free extensible model for Siri. They should have, because people are on aggregate more creative than any big company including Apple -- thanks to survivorship bias.

They did, its just not well advertised. Any app on iOS can expose actions to Siri, for example the BBC Sounds app exposes actions to start playing a specific radio station. Those actions are then available from Siri on your phone, and any HomePods that have been granted access to execute actions on your phone.

On your point of looking at it like a command line, those same actions are also exposed to Shortcuts, allowing you to compose them in arbitrary pipelines. The output of one action can be passed into the output of another or even stuffed into a variable for later use.

I've developed for all three of the major voice assistants and Siri is by far the best simply because it completely throws out the window the concept of developing a "voice application", voice is just one possible entry point into an iOS app.


Interesting take. I am kinda excited about voice assistants and the command-line style use cases for them, too, except the closed nature of Siri and the dominance of Apple has probably set them back 10 or 20 years.

Yes, if Siri was open and any iOS app could easily make a "voice CLI" to their apps that would awesome. But, that's not the case at all.

I am a heavy user of Siri, too, with an Apple Watch and a HomePod in every room of my house. It's better than nothing. But it is still objectively terrible. It can barely do more than set timers, send text messages, and control the lights and music.

That's not useless, but it's not even close to what I wish I could do with my voice (which is basically what you describe).


You can do it using https://developer.apple.com/documentation/speech or is it actually missing anything necessary for a “voice cli” app?


I haven't actually looked at in depth for a few OS releases (I was excited to make a voice CLI I could use with my watch, since I commuted by bike).

I think they may have made some prgress since then, but I don't think the Speech.framework you linked to provides what we are talking about when we say a "CLI for voice". I think that API is only useful when you've already opened the app and it is able to record live audio (or process previously recorded audio).

So you can't just be like, running from the feds and shout to your watch or phone, "Hey Siri, destroy evidence, password 'Rosebud'!" and have your arson app torch your meth lab.


Hahaha! Nice one :D

And here I am thinking about mundane applications such as: "Hey Siri, limit order 10 shares of ACME for $44.95"

or "Hey Siri, call John Doe via WhatsApp"

or "Hey Siri, invite John Doe and Jane Goodall to bowling on this Sunday 7PM via someSpecialCalendarAppThatIsNotTheStockCalendarApp"


"Thank you, I've ordered 10 shares of AYME for $4495"

The fundamental problem with voice assistants is an utter lack of common sense. Any human assistant not completely asleep at the wheel would flag the above because it's outside your normal patterns and ask for clarification.

Voice assistants will just go ahead and do it. And then play a Siberian Metal Polka song because they thought they heard that too.


Ironically enough given its reputation, but this is what Bixby is great at - controlling the _device_ with your voice using short commands - "Open Camera", "Take photo", "Delete last photo" etc all work as expected (as a toy example, far more complicated stuff is possible even before including their Siri Shortcut equivalent)


Siri sucks because it does t understand what you are saying at all, not because there aren't a million things hooked into it.

To be fair, all the assistants are extremely bad at it, Siri just seems to be at the bottom of a bad list.


Well, looking at the skills offers by alexa, I am not sure if a siri appstore would be worthwhile. 99.9% of alexa skills are totally and utterly useless.


Can you not use voice typing with swiftkey (or any other keyboard) and open termux? I've done that a few times on android


I've loved the idea of a verbal commandline, too, especially since watching the Play For Tomorrow episode "Shades[*]." It's a teen brave new worldish sort of thing, but all of the characters are constantly interacting with databases, networks and media (including VR) with this very stylized language, not how you would talk to a person.

That's the problem, really. You have to use a stylized language on a commandline. A GUI is really just giving you a graphical diagram and guide to a really stylized language (if you reduce a GUI to its clickable/typeable areas.) It's very hard to draw an arrow pointing at a picture representing a particular predicate in the very restricted bandwidth of serialized sound.

The reason why "Shades" was so interesting to me was that its future teens clearly all knew the stylized language well, and obviously would have learned it in school. To have that kind of universal training in a certain form of computer interaction would require a universal OS, like what UNIX quickly became.

So maybe this is a conclusion coming off the above at an odd angle, but what I'm actually saying is that commandlines are hard and people don't like them. Visual, and even tactile stuff have the ability to offer so much feedback and so many affordances for amateur operators in comparison to audio that audio can't compete. Audio feedback is all exposition and nothing to hold onto or see. And just like film exposition, it often comes too slowly to keep people's attention, or too quickly for people to understand. It's always going to be easier to grab the knob and turn it, or to find the button and click it.

The major advantage of audio interfaces is that they're cheap both materially (you don't need a screen or a control panel) and for the user (you don't have to stop walking, or sit down, or grab a thing, etc. you just speak.) For me these advantages are actually dominant over innate usability, which is why I also believe with you in the huge potential of audio interfaces as a platform.

I just think that the only way to get there is to build a universal and open verbal protocol, with standardized interfaces for accessing databases, media, and networks. Something that children can (will) learn, and will have a lifetime of use. An opinionated and efficient design moonshot like the web (or the Bourne Shell or Gnu Corutils) turned out to be. A platform whose innovation comes from addition of new ways and not from subtraction of the old ones. A platform that holds onto backwards compatibility as long as is practical, thinking in terms of how many decades to keep things around, not how many releases, or possibly even a closed kernel that will never change built on top of some mathy orthogonal basis system.

* I wish this hadn't been deleted from youtube.


I suspect that most (or all) languages are too verbose for command-line-like voice interfaces. As capabilities and complexity grows (imagine ffmpeg), it’d be a major issue.

I recall there was a post here by someone who used voice to write code. They had to end up with a lot of vocal “shortcuts” for various bits, that ended up making their speech much less legible than a normal “query” language.


> I wish this hadn't been deleted from youtube.

Is this the series?

https://www.youtube.com/watch?v=7oFhdQJxhmc



I'm averse to any kind of voice assistant. I'm the kind of person who just finds it faster to type in my query or go hunting for the information I want directly. I also come from a time when voice recognition used to be just barely functional, so I'm biased to think the phone will mishear me. I also have a bias towards thinking that computers aren't meant for natural language questions - I'm still very much a keyword query maker. To this day, I have never asked Siri anything, nor Alexa, Cortana, or whatever Google is calling theirs right now.

I know I can't be the only one, right?


I find Siri great...while driving(to call, play music, change volume and ask for news). And that's the only use I find for it.


No. Even asking the weather, I can get a weather report with far more context in about 30 seconds on weather underground either on my phone or on a computer.

And for places where I can't just go to a computer or phone, e.g. my car, I find Siri frustratingly difficult to instruct to perform specific tasks, especially without taking my eyes off the road or hands off the wheel. Even setting a GPS destination by voice alone is hit or miss much less finding a podcast that hasn't been pre-loaded.


I stutter, and as a result just don't like talking. I'll never use an interface like this.


I wonder if there was an interface that required singing, or possibly it would play classical music while you talk - would it work for you?

I say this because of two things I've heard (possibly misconceptions)

One was that people who stutter can sing ok

the other was the movie "the king's speech" where playing classical music while talking would prevent the king from stuttering.


As I got older, the stuttering became less and less of a problem; now in my late 40s it happens rarely. I probably wouldn't actually stutter talking to a device (might be different with other people present).

I just that since I was young I have always preferred text over voice whenever there was a choice, and this is no different.

I think the singing thing is true but someone who is extremely self conscious of how their speech comes across is never going to sing instead.


I was like this up until I got my first Apple Watch.

Maybe because it’s that I’m white and speak english, but it’s dictation is unbelievable and I can even send Facebook messages or texts an average 9/10 times without needing to modify it.

I’ve never - for instance - had an issue where I’ve said ‘set a timer for 5min’ and it’s never worked.

I think everyone’s mileage varies.


For me it's that my brain just never got used to asking the assistant to do things. Sure I asked it the weather or for the news when I first got it. But after a while I noticed that if I wanted to know these things I would just pull it up on my computer or phone without even thinking to ask the assistant. I'm sure if I used it longer I would get used to it.


Outside of cooking, I'm with you. I use Siri for setting timers while I have my hands full in the kitchen - and even then - it only allows _one timer at a time_. My phone has multiple processor cores, but it requires the assist of my 15 year old microwave to track both rice -and- veggies...


I feel the pain with setting an alarm... I have my phone set to 24 hour time. When I say 16:04, it's distinctly different than 04:04. Yet, I tell Siri to set an alarm for 09:40, it sets one for 21:40 like an idiot.


In some countries that use 24 hour time notation it's really uncommon to use that in spoken language[0]. When I say four past four in Dutch, it could be 04:04 or 16:04. So even if you've set the clock to 24 hours, that doesn't mean you would actually say sixteen when referring to 16:xx.

You can set the region in the settings, maybe that influences the way Siri interprets time? (I would be surprised if it does, but you never know)

[0]https://en.wikipedia.org/wiki/Date_and_time_representation_b...


Would you never use the 24h format when speaking? E.g I’m Swiss and I would often say 16:04 when referring to a precise time like a train departure. But four, quarter past four, half five and quarter to five.


No, I would use: four (vier uur, so literally translated four hour), four past four, quarter past four, half five, quarter to five.


It's pretty rare in North American English at least. You almost never hear it and it automatically brings up a military vibe to me.


I don't have a ton of experience using Siri, but it seems like it might have issues with time in general. I got an email just yesterday with an invite to a Zoom meeting attached, and I was about to just import the ics file into my Apple Calendar but then I noticed the email had a line saying Siri had detected an event and could do it automatically, which it did, but the listing was 5 hours early and scared my wife when she saw it because she thought I had scheduled something for the middle of the day when she was at work.

I have a lot of trouble understanding how plain text in the email saying 6:00 PM Central Time could have been interpreted as anything other than 6 PM Central Time.

It was also annoying that you apparently can't edit a calendar entry. I had to go to DuckDuckGo to search for how to change the time and you need to click and drag the box containing the event. There is no UI to just type in a time. I guess they did this to be able to deploy an identical app on Mac OS and iOS? Neutering desktop applications so they work better on mobile is an increasingly annoying trend.


I face the same annoying issues with Siri when I’m trying to set reminders from the Apple Watch. I often have to repeat 2-3 times then remove the incorrect ones.

But you have worst IMHO in other languages! When speaking German units are spoken before the tenths (fünf-und-dreizig to say 35). That completely confuses Siri when saying dates, hours/minutes, or just big numbers.


Dreissig*

I speak German to Siri everyday and she never has this issue with setting alarms. Maybe it’s a dialectical difference?


I’m not native and would be lying if I would say my German is great. But I don’t think it’s just about my prononciation, I see it often with German auto-generated subtitles or transcriptions, that dates or big numbers are transcribed incorrectly due to the unit-before-tenths grammatical construct.


A minute off from the 09:41 of Apple ads and screenshots:

https://www.businessinsider.com/heres-why-the-time-is-always...


Workaround: say hours at the end. I just said "hey Siri, set a reminder for oh nine fourty hours" and it handled the time perfectly by setting it to 9:40 AM.


You’re holding it wrong — How to blame your users

https://uxdesign.cc/youre-holding-it-wrong-how-to-blame-the-...


Suggesting a workaround for a failure of the software is not at all like suggesting they're doing something wrong.


I like Apple in general, but using Siri is the most frustrating experience Apple has ever rolled out to the public. It's absolutely awful and useless in its current state for anything but setting a timer or alarm.

Google Assistant is several galaxies above it in every aspect.


Wow, opinions vary here a LOT.

I find Siri very capable and useful. It's surprising to me when I don't get what I want out of it.


Kind of related, but has anyone else noticed iOS text autocorrect has gotten notably worse over the years?


I believe at some point in the last couple of years (iOS 13 maybe?) they switched to a machine learning-based system and yes, it's been significantly worse since.

I cannot put into words just how frustrating it is to have the device change a correctly spelled, correctly chosen word because it thinks I may have meant something else. Double points if I backspace to change it back and it re-"corrects" me again.


> Double points if I backspace to change it back and it re-"corrects" me again.

That's just awful design. Another quirk is that it can autocorrect the last word you typed when you press "send". Sometimes I proofread myself and still end up bitten by the autocorrect because it changed something when I pressed "send".

On Android when you backspace once after an autocorrect it reverts the autocorrect. It's not great but still better than what iOS does.


This happens to me, too.

It’s particularly annoying when writing a tweet. I’ll correct a misspelling and — presumably as the text field loses focus when I tap on the “Tweet” button — it’ll change the word back to the misspelled version and use that for the tweet text.


Love it when it replaces a correctly chosen word while swipe-typing the next few ones because it would fit better in the new context.

Boggles my mind how you can let something like that into prod. The things it leads to are

- Errors that make the message hard to understand that are (here it replaced “that” with “they and “here” with “get”) only caught while proof-reading or not at all

- Having to go back and re-type 3 words for correction.

I check the word I just swiped for misrecognition, not one that was previously correctly recognized, god damn it.

The actual success rate of that piece of garbage is maybe 5%.

That said: Of course I’m happy they have a swipe feature at all, albeit significantly worse than Google’s (about on par with Microsoft’s). But you don’t have to make the feature intentionally worse, do you?


The re-correction thing is unacceptable, imho. No predictive text algorithm will be perfect, but why in holy hell would they let it ignore explicit correction is beyond me.


It just refuses to learn!

This is inarguably worse than what we had with T9 like 20 years ago, all running without cloud support on a limiting embedded platform.


Yes https://discussions.apple.com/thread/252621788?answerId=2554...

If you’re wondering why “Access is denied” then that should inform you about how Apple deal with negative customer feedback these days.


I always turn off this feature in both Android / iDevices because it can act like a keylogger - the Terms of Services for this allow both Google and Apple to "collect data to improve their services".


I switched from Android to iPhone a few months ago. The hardware is much better but my god does the phone feel dumb. Ok Google is leagues better than Siri and I miss the hell out of certain Android features like better spam detection and the on-hold service.


to hell if I will use a phone again without the on-hold service. absolute god send when you have to call any corporation in modern times.


As the article points out, it seems apple missed the ball with integration into other apps. I'm not completely certain, but I don't think apps have the capability to "hook-in" to Siri in any useful way. Apple's obsession with maintaining control means they never nurtured this kind of development.


There is app integration with Siri, via tasks and intents. There's also Siri Shortcuts, which can integrate with custom intents.

https://developer.apple.com/design/human-interface-guideline...

https://developer.apple.com/documentation/sirikit/adding_use...


Ah, thanks for the info.


Actually they do, for the most use cases. Sending messages, performing certain actions like switching things are the biggest use cases for most 3rd party apps for regular users. For the rest (power users), adding a Siri shortcut within the app almost covers every possible scenario.


It exists now. You don’t even have to say “on Spotify” for every song anymore (holy shit!!!!). They’re certainly moving late, but they’re moving.


I still have zero desire to speak to any computer with my voice. Not saying it's a bad product feature. Probably just a psychological profile some of us fall into.

EDIT: Yes I'm already aware it can be useful/quick for doing X :)


I got a free Google Home puck with some purchase a while ago. The only thing I've used it for is setting timers in the kitchen, which it's quite well suited for as it's genuinely useful to be able to do this hands-free, and to have multiple timers going at once. None of the other things it can do are of interest.


From what I've heard about these products since their introduction, setting alarms is the one thing that they're unambiguously useful for.

edit: For that they don't need to be networked, of course.


They are also handy at dimming/turning lights on/off or starting Spotify playlists while in the shower. Nothing essential TBH.

[Edit] Google Home seems to be pretty good at separating your voice from background noises, like shower/TV.


Works well for playing radio stations too. I also use my Echo to control my FireTV. Nice to be able to pause and grab something without hunting for the remote.


The speech recognition is performed in the cloud, not locally, so under the current architecture they do indeed need to upload the audio capture following the wakeword.

One of the other things I used them for frequently before disabling it on all my devices and removing all of the networked mics from my house was things like "remind me to $THING at 21 tonight".


Fwiw, starting in iOS 15, Siri will process speech on-device, which is huge.


I hope this doesn't make Siri's quality even worse, but it seems inevitable.


It's other on-device processing that's planned in iOS 15 (such as scanning my local files without consent to report on me to the police) which is going to keep me from using devices with it.


It's much slower even with the A12's dedicated cores for this purpose, but I assume that'll improve over time.


Try this - turn on airplane mode and go into the notes app and do a dictation.

On the iPhone, the speech to text can be done on the phone. I'm not sure if this is the case for Siri, but it doesn't need to be done in the cloud.

That said, the words to intent mapping doesn't appear to be something that the on phone model is able to handle.


> The speech recognition is performed in the cloud, not locally, so under the current architecture they do indeed need to upload the audio capture following the wakeword.

I bet it wouldn't have to be if its vocabulary were restricted to timer related words.


It doesn't have to be, but it makes the device cheaper and they're going to gobble your data up in the cloud anyway.


Same thing for me with my Alexa devices. They're a good kitchen timer and nice speakers, but everything else is not so great. Often they don't understand me at all and they need to be rebooted once a month, because they'll hang themselves up.


Ha ha that's all we use our Alexa for. I really should just get rid of it. I once set it to turn our smart bulbs red and play Star Trek red alert sounds. That was fun for 5 minutes.


Same. Using it to add things to my shopping list is useful too.


It feels like unnecessarily complicated way to give instructions to you phone or computer. It's not that I can't see it being useful in some cases, but more often than not, it gets in the way, if it's not something you use regularly. I just want to be able to turn off Siri permanently. Just remove all possibilities of me triggering Siri. Why Siri is included on the Mac at all is something I don't understand.


It complements the UI by summarizing properly: “Will it rain tonight?” and “Set an alarm for 6am” are the only two usecases I’ve been satisfied with, but they work much better than the UI.

I’m not sure it’s worth the development time. But it’s like: Let’s create a UI with all info we think the user wants, and everything else will be done by Siri.


But we asked Siri if it was going to rain, it said unequivocally no, we didn’t take umbrellas or raincoats and it rained. We got soaked :(


This isn't inherent to voice assistants though (which is what this subthread is about). Siri is a known, particularly poor implementation of a voice assistant. I don't have an Alexa, but I just moved to a city where rainy days come out of nowhere. My Google Home is concise and precise about the chance that I'll need an umbrella, including what time the chance of rain starts.


I don’t know if it’s related but the weather app on iOS frequently takes minutes to update the weather for my location even when looking directly at the app. It’s incredibly annoying and has led me to believe the weather was going to be completely different more than once.


> just want to be able to turn off Siri permanently.

It's your lucky day! There is an "Enable Ask Siri" checkbox in System Preferences on macOS, and a similar toggle in the iOS Settings app.


Yes, you think that would do it, but no. Now instead of triggering Siri, you now trigger "Do you want to enable Siri?". I never want to press something that might trigger anything Siri related.

I want a check box that says: "Ask to enable Siri", which I can uncheck. So "Enable Ask Siri" should be off, and "Ask to enable Siri" also off.


I see, actually I do remember macOS used to occasionally nag me to enable Siri (in response to accidentally pressing something, I think). But I don’t think it’s done that since upgrading to Big Sur.


I want the same, but for Android. It is possible to uninstall it, but it always come back with the next upgrade.


I find it works well for a few use cases:

- play some song or playlist

- open all the blinds, or a set of blinds, or all lights, or some lights. And close then close them/turn them off.

That’s about it. The only reason it works well for those use cases is because the equivalent menu/physical controls take too much time to use vs voice commands.


Also it’s really useful if you are in a car - getting directions, calling people, texting etc without touching your phone.

Sure lots of cars have this already with voice, but Siri et al are much more reliable at it.


Meh, the ~90% success rate kills that purpose for me.


Me neither, but speaking to my watch when I’m on my bike, waiting at a traffic light, works very well! I’ve never seen a use for Siri on my phone or laptop but use it often when biking around to set reminders or timers without losing focus over my environment.


Same here. I'm curious if it has any correlation to having an internal voice. Do you have an internal voice for your thoughts?


I don't really have a binary answer for that. Sometimes yes, depends on the type of thinking being done.


Wait, don't all people have that? If not, then how do they think?


I don't. I know a friend who doesn't as well. Both of us mainly think through an awareness of some kind of graph structure (non-visual).

But what's more interesting is that through talking with others we've discovered one interesting thing about what it's like for other people to hear a voice in their heads:

In most people's dreams, when someone talks, they can hear the voice, which has a specific pitch, rasp, etc. For me and my friend, in our dreams we can never characterize the voice. We just become aware of the sentence spoken and the time passed (even though in a way it feels immediate), but if we focus, it's clear that we cannot recall any sensation of hearing a voice in any way. It blew our minds off when we realized that other people's dreams are different in such a way.

I can think of sentences. But they are not spoken in my head. There is no voice, which would be male or female etc. And I only do that when the sentences are the goal. They are not the means.


> In most people's dreams, when someone talks, they can hear the voice, which has a specific pitch, rasp, etc. For me and my friend, in our dreams we can never characterize the voice.

I’m not sure that this is true. I think in terms of my dreams they are closer to what you describe as your experience; mostly visual, with speech something that’s more often understood than heard. I imagine most people have something similar, and not so analogous to real life.

I do have an inner monologue in my waking hours though, and it’s very interesting to think about what it would be like to not have one!


I’m the same. When I’m reading, it’s more like a silent with blurred shapes than me speaking in my head. A few friends say they can’t read subtitles while watching a movie. For me, they’re just an overlay to the scene until I grasp the meaning, kinda like you realized: This is a square and this is a circle. It makes reading more enjoyable as it feels like you’re in an holographic projection, but that can make some scenes quite disturbing.


That is amazing, I had no idea. It's not something I ever thought about, I just assumed that everyone thinks in the same way.


I have a strong internal voice and wondered what is was like to not have one. People describe reading and not hearing the voice and this sounded strange to me.

Then I noticed that I don't think with my internal voice when coding. I'm still thinking but not with sounds in my head. That blew my mind and made it easier for me to understand other's internal experience.

I wonder if people who rarely have an internal voice are less interested in things like podcasts, where the information in conveyed aurally.


I don’t have an internal voice. I mostly think in concepts or shapes.


I genuinely dont believe people who say this. You cant think in shapes when thinking about your personal relationships.


(Not the commenter you replied to.)

That's correct. But neither can you think in sentences in that case, unless your relationships are just calculated means to an end. Whether you want to continue a relationship is not a conclusion of a sentence, but something you know without words, just like you know without words whether you want to eat the dish that's on your plate.


I'm the same, but mostly in the following senses:

- My fingers will more reliably get me to where I want to go in a reasonable amount of time.

- I don't desire for The Google to know the instant I turn my lights on or possibly take a dump.

- It's only a matter of time before "always on" turns into "always listening".


Siri and similar voice assistants work really well as an universal remote control.

Right now I can ask Alexa to open or close the curtains, to start cleaning my house, or to play some music or podcast — all without having to find where my cellphone is or risking getting trapped a couple of hours on the internet.


I just think I want a voice based CLI more than some NLP assistant for specific situations.


The thing I really want with Siri is better user documentation. Give me a dictionary of all the things I can say to it! I'm only going to guess the obvious ones and then sometimes try a few experiments, but lack of awareness of what it can do means I just do the handful of things I have seen it do successfully.


That will probably cause it to lose a feeling of magic for the user. If the list of supported phrases is listed out, then you'll see it for what it is: a voice based CLI.


but I WANT it to be a voice-based CLI! which, yeah, I recognize that most users don't want.


this is generally a complaint i have with most modern software.

With the older versions of windows like 95 and 98 (that's the oldest experience i have), i read the manuals, tried things and slowly learned the intricacies of the OS.

Modern smartphones, tablets and computers seem to come with increasingly sparse manuals, let alone documentation. Take iOS for example - i have no idea how to manage windows on my ipad, i don't know how to close them, window manipulation is not obvious and so on.


Apple would improve their ecosystem if they let the consumers they supposedly love choose a default voice assistant.

Siri is so bad compared to Alexa, I laugh out loud when I occasionally try to use it and it has no idea what is going on.

This would be a boom for competition and the entire ecosystem.

Will Apple’s ego allow it to be bruised like this? No.


If they can guarantee my data never leaves the device I might want this. Otherwise, hard pass.


For people that don’t expect the entire internet to be stored on their device, it should be an option.


Digital assistants are the height of sloth. Older technological inventions were truly transformative. Prior to the intention of the washing machine people wasted hours a week doing moderately hard physical labor. The washing machine of course still requires that somebody be present to change the loads, but otherwise people are free to do more or less anything else while the load is running. This was truly transformative.

What does a digital assistant get you? It possibly saves you a few seconds, or allows you to set a task while walking or driving rather than forcing you to remember the task all by yourself and then later use the phone with your hands when the time is available. This does not save you time, instead it only saves you impulse control. ie, the task takes the same amount of time, but now feels like it requires no declarative action on the part of the user.

It's not even clear that voice commands are better than using a GUI. If I'm unfamiliar with a GUI, I can look visibly for cues which point to the function I'd like to use. Voice commands require me to memorize the commands and syntax which might be used. It's more of a "simplified command line" than it is a GUI.

And then of course, n% of the time, a perfectly-formatted voice command fails for invisible reasons. And so n% of the time it does not save you time.


that’s a valid criticism, that digital assistants don’t bring much time savings, but they’re also a necessary stepping stone to a (far?) future where we talk naturally with our computers rather than type/tap (transformation via incrementalism). beyond the phone, apple clearly sees siri as a necessary component of its glasses, home, and car initiatives.


I'm still not convinced that any of the digital assistants have their job figured out. They are useful when driving, or if you want to listen to music when cooking. Beyond that I fail to see an actual use case. They could be a major benefit to disabled people I suppose.


That's where Siri absolutely and completely fails for me. It doesn't understand 90% of the song or band names I ask to play. Maybe it's good when all you ask it to play is "The Beatles" or "Kanye West", but anything more ambiguous than that, it will produce a wrong result. Google Assistant is way better at nuances.


I am impressed that it gets Khruangbin right, but I'm guessing that apple feeds Siri with artists that are popular with their target market.


I felt like it was really helping me when we had kids, and I was like, holding a baby a lot of the time. But that is form of temporary disability, maybe.


Pretty much the only thing I use it for is setting a timer or alarm. For anything more complex than that it always seems easier to just use the app.


Siri the voice assistant is disappointing, but I find Google's and Amazon's to be disappointing also and I think it's an inherent problem of voice assistants and AI. If Google Assistant can get basic tasks correct more often, I just end up irritated that it can't get more complex tasks correct.

But Apple's been branding some simpler heuristics as Siri and those are excellent: - Suggesting data to add to a contact based on scraping from elsewhere - Associating a phone number that's calling me with a person who emailed me with a phone number in the signature - Suggesting apps to open based on time, location, connected peripherals, or recent activity - Suggesting driving directions based on calendar items or history

None of these make me say "wow! AI!" But they do make me feel like my phone "just works" which is even better.


I believe this article conflates two separate pieces of voice AI: speech-to-text and processing the text.

Google’s speech to text is amazing, possibly related to the “less privacy” part. Siri can handle slightly more complicated commands but so many times I see it transcribe what I’m saying… then in a blink change all the words.


This is not a diss on Apple or their employee, only an example of a common fallacy. But, I was in an online meeting with an Apple dev. Someone was giving a presentation. After the presentation that Apple dev said "There were zero mistakes in the captions, your English was so clear" and my first thought was that thinking there is a "correct English" is why these systems don't work. For myself, it took me a long time to stop thinking of English with different pronunciations and usages as wrong. For example Singapore. They are not speaking incorrect English. They are speaking their form of English which is fully valid and they have no problem communicating with each other. It's tech's job to understand all of it, not to get everyone to speak "California news broadcast English"


Using siri on my iPhone is 'okay' for really simple things, like setting timers, asking for the weather. Sometimes setting a reminder works.

But I also have a Homepod in my livingroom and using Siri on the Homepod is really useless, it just feels abandonded. The worst is when Homepod handles my Siri commands, which always happens because the iPhone sends the requests there, except in random occurences. Then it fails almost always ("can't do this on Homepod") but then there is no way I can force it to handle it on my iPhone.

Madness. I don't know who at Apple actually uses this.


Google Home can hear me whispering on the other side of the room, and usually do the right thing. Meanwhile, Siri on my watch gets triggered by the hiss of a running tap or a sizzling frying pan.


My experience has been that Google Assistant has always been better than Siri because Siri does all NLP on-device and Google Assistant offloads this to the cloud, where it can improve over time due to ML. Nonetheless, both are horrible in their own ways, and neither is very good from a societal perspective. I've taken the stance that voice assistances are fundamentally bad because they teach people that it's acceptable to carry on conversation with a computer that is likely being used to spy on them, either directly or indirectly.

I don't allow voice assistant products in my house, and on devices where it's non-optional (mobile phones), they are turned off on all devices in my house. I also personally find it absurd that there would be anything /so/ important you can't wait a handful of seconds to use your hands to do with higher fidelity. Everyone always talks about using maps or something in your car... you shouldn't be changing destination while you're driving anyway, you should have a destination in mind (or not) and be focused entirely on driving with both hands on the wheel. If you need to futz with your GPS, pull over, whether you're talking to it or using your hands.


What's the difference from a surveillance perspective if you say "hey google, what's the weather today" vs typing "today's weather" into Google? If the assistant is processing the language on device, is there any difference at all?

> I also personally find it absurd that there would be anything /so/ important you can't wait a handful of seconds to use your hands to do with higher fidelity.

I think people often choose a voice assistance because the opposite is true - the tasks they send to their voice assistant are very unimportant. They are playing a song, starting a timer, or finding out what time a baseball game starts.


Taking a slightly broader perspective of Siri as a simply a voice interface to the device, it has been incredible tool for two of my kids who are not neurotypical. For some who are differently abled, it’s not really about the convenience of saving a few seconds, it’s about being able to do something otherwise not easily done.


Smart lights. Music. Timers.

Changing the brightness level using an app sucks. You have to dig out your phone, unlock it, open the app or go the widget, find the light you want to dim, click on it and then slide the brightness to where you want it. With a voice assistant like Alexa I can just tell it what light or room to dim and what level to dim. It's WAY WAY faster to say "Alexa, dim the kitchen to 50%" than to go through the app.

Similarly with music or timers. It's just way easier to tell Alexa to play that radio station or album or song or to set a 10 minute timer than dig through smartphone apps and deal with hooking up speakers somehow.

Of course it COULD spy on me, but so could any of my electronics with microphones and a net connection which is most of them these days. What's stopping your smartphone from spying on you or your laptop even with the voice assistant disabled?


There's probably scores of Apple employees working on Siri week-in, week-out.

But in all the time I've had it, I use it for literally two commands, both of which I use often:

1. "Set timer for x minutes" 2. "What song is this?"

Using it for anything else bubbles up too many false positives to make the experience worthwhile.


I can count on my hand the numbers of times I intentionally wanted to use Siri. 98% of my interactions with Siri were accidental to the point that I ended up disabling Siri until I wanted to use CarPlay. It's just not good and it doesn't serve me any useful purpose.


About 80% of my Siri use is setting kitchen timers, 15% setting alarms, 5% setting reminders. It's really good at the first two things and adequate for the last. I don't trust it for anything else.


My personal experience has been that on a few occasions I’ve tried things like “navigate to 123 oak street, Harrisburg Pennsylvania” or “play ‘all you need is love’ by the Beatles” or “dial John smith”. And so far Siri has a zero percent success rate with any query no matter how simple. So I gave up trying. Really those 3 things are all I really want, because the only time I need Siri is when my hands are busy because I’m driving. And the 3 things I want to do are: navigate somewhere, play music, or call someone. That should be any digital assistants bread and butter. Siri just can’t seem to do any of those things.


If you have two contacts with the same name and you ask siri to call or add them to a calendar event she'll ask you which one. And won't be smart enough to understand when you say first or second.


I just got my first iPhone some months ago. I really wanted to give Siri a chance because of privacy. But more often than being useful it just bothered me during video calls when it suddenly started to speak. So I turned it off again. I would like to have it on when I'm using Apple maps for example, my kids would have fun asking Google stuff in the car. But that doesn't seem to be an option.

Edit: I know I can put Siri behind a button too, maybe I can give that another chance.


Ten years ago I hated the fact that they gave these assistants a female name and voice. Today I hate the fact that after 10 years they still have not changed this. At least Google didn't give theirs a stupid female name, but they still use a female voice.

It's not hard. Give this software a neutral name and a neutral voice: https://www.youtube.com/watch?v=kRauhbZqJCY


You've been able to change the voice in Siri for years under settings

https://www.macrumors.com/how-to/change-siris-voice-on-iphon...


“What we really want to do is just talk to our device,”

If we could "just talk to our device", I'd be able to ask it to change her name to whatever I want.

But, yeah no. I don't _want_ to really "just talk to my device". Years ago, simple things just worked (i'm on android)... I could just say "e.t.a." to google maps and it would say the eta to my destination. Now it just gets confused or starts talking about the wiki page or some such nonsense.


"the job" is different for everyone. I've accepted that siri can do reminders, conversion, timers, and simple HomeKit commands pretty well.

I've never seen a more complicated tech demo of Siri that implies I should get more out of it, so saying it fails broadly in any way is meaningless. (edit: typo)

It's a piece of software on your phone you effectively get "for free" that you can completely ignore if you so choose.


My problem with these assistants is that you can ignore them but not truly disable them or remove them.

Cortana is always waiting for me to talk on Windows. Google Assistant is always waiting to hear "Ok Google" on my phone.

I never want to talk to my computers, ever. I wish I could remove this functionality entirely.


Am I the only one who thinks Siri "just works" as it should?

Sure it's far from perfect, but it almost never failed me to call or text someone, ask about weather, take a note/reminder, switch the lights using my smart outlet, or to set an alarm (my personal highest use-case).

Sure it could do more, but I don't thik it didn't "figure out the job".


I got my parents a Homepod and they kind of hate Siri but also ruefully love it like one has to chuckle at the dog that ate your shoes.

They're huge classical music nerds, and I've watched them, in vain, recite to Siri over and over and over the desired concerto, orchestra, etc – trying to find the right incantation to get their Homepod to play the music they want. Usually they just get fed up and play it with their phone.

My dad likewise chuckles when Siri inevitably gets his spoken grocery list wrong. One of his favorite games is getting people in earshot to work with him to figure out the correct pronunciation of words for Siri to understand what they're saying.

I think how people accept Siri as a tool is by finding the few things that she does perfectly and never doing anything but those few things. Ever.

Anyway, here's a scattershot of some stuff I just tried that Siri fell on her face over (on iOS 15):

* "Send a text to Katrina, Mom, and Dad that 'I don't need their help'" > Creates text to Katrina: "Mom and dad that I don't need her help"

* "Show me photos I took last month" > Simply opens Photos

* "Show me photos I took in Scotland" > Simply opens Photos

* "What are the hours of the closest grocery store?" > shows assorted list of nearby stores, no hours shown

* "What is 13 modulo 5?" > 13% = 0.13 [clearly the Calculator app doesn't support modulo operations but still: at least say that?]

* "Which novels won the Hugo last year" > No response, link to Wikipedia for Hugo Awards and an article on "Novels that won both the hugo and nebula awards"

* "What time is low tide today?" > no response, link to various tide times websites

* "When is the next Jon Stewart episode coming out?" [an Apple TV property] > no response, shows list of websites talking about Jon Stewart


Some of my annoyances:

* What’s the weather for tomorrow? > I can’t see the weather that far out but here’s the weather for the next ten days.”

* “When will it stop raining?” > It looks like it’s raining right now.


Even though I like Siri this reminded me of this one time I was driving with my Apple-hating friend with me. We were discussing whether it'd rain, I asked Siri, she said it doesn't look like it's gonna rain today.

Literally 10 seconds later, it started to rain, giving my Apple-hating friend enough material.


Yup, this list also fail at my side, but frankly I've never tried them before either. I think you're right about accepting it as a tool, I figured out what she does perfectly within my needs, and always ask things within that domain.


I've asked what time it is in PST once. It told me the current time in my timezone, with the PST timezone qualifier.


Wow, this is a great example of how crap Siri is. Let's do this – all of these work perfectly with Google Assistant:

"Time in PST" gives me 3 links. The first 404's, the second has no clock but alerts "No locations currently on PST" (which is true, its PDT right now – but come on), and the 3rd has no clock but prompts me to install a "FREE iOS app." Fail, big time.

"What time is it in PST" gets me a web snippet showing "PST is 8 hours behind UTC." Fail.

"What's the current time in Pacific Standard Time" runs it through Wolfram, giving me a LOT of unrelated info but admittedly also a clock showing the correct time. Success... sort of. Wouldn't help while driving.

"What's the current time in PST" gives me "Sorry, I can't help with that, but you can ask me the time for a specific city." Fail.

"What's the current time in Los Angeles" correctly shows me a Siri prompt with the time. Success.


I'd never think of asking it what the time in my particular city is. I reckon that's how "normies" think time works, and Apple has tuned Siri to their needs? As opposed to the needs of people who may have some knowledge about nerd things like time zones?


Sure, but conference calls usually come with a timezone and time, not a city (which isn’t exactly unique to a timezone)


on most interfaces, timezones usually specify a region or a few cities as examples.


It clearly means that you, as a normal user, shouldn't care about time zones. It's an implementation detail you should not be aware. What are you doing, asking about specific timezones?


Setting alarms and timers is the only thing where Siri works for me more than it doesn't.

Asking it to play a certain song or band or anything more than a 3 word command will fail or result in long stalling with an eventual "sorry didn't get that" reply. This is on WiFi.


You are probably talking to Siri like you're talking to a computer. However, most people talk to Siri like they're talking to a human being.


Yes and no. As a computer engineer I'm probably biased about what it can and can't do technically, so I ask things that it can do. From a user perspective, I use a very natural language instead of robotic/feelingless commands and it just works anyway.


Take a note? I wish.

Often I'll think of something while out cycling that I want to remember later. So I try to tell Siri using my watch. Siri just tells me "Sorry, I can't help you with notes on Apple Watch".

But it will let me send myself a message. So I do that instead. I'm not convinced iMessage was meant as a note-taking platform, but hey, maybe Apple have plans.


Do people really need these assistants that badly? I use siri to text and call in my vehicle occasionally and that's it. I wouldn't really care if phones came with no assistant. In fact, I think I'd prefer it.


I have a couple of good things to say about Bank of Americas virtual assistant Erica. It’s very niche but I’ve always found relevant information when I use it.


The assistant I’ve been really impressed with was Alexa. I remember asking it all kinds of complex things and it just knowing the answer.

Google seems to be somewhere in between.


Is there an objective benchmark to measure the performance of voice assistants? Usually, NLP tasks have ways to measure their performance.


I use Siri several times a week on my computer to find my iPhone. And that’s about the only thing I use it for, but it’s super useful.


I find it pretty useful to set up reminders when working without having to unlock the phone or even pick it up.


what is the current state of the art in open CSR/transcription? it feels to me that theres a lot of potential in these assistants that seems locked ip by the big players


My favourite Siri bug: If it fails to understand the name of the person I want to call, it suggests to call myself! "OK, so you want to call <your name>?" This bug exists since about 10 years, and apparently has never been fixed.


Did you try answering "yes"?


Yes, believe it or not, it will try to call itself, and fail.


its really terrible, apple should just delete siri, its too embarrassing


Open Siri and say "hey google, what's amazon's stock price?"




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: