Why isn't iOS better at data recognition?

[cameronjpu]

OK I've waited several years in hopes of some improvement. I don't feel like it's much to ask.

Apple knows if a phone number is an iPhone (because of iMessage compatibility) - why, when I add a contact directly from an email, does it assume the phone number is a home number instead of querying it and labeling it mobile if it's another iPhone number? And why when I add a contact in that way, do I FIRST have to add the name and email address, and only THEN do the data detectors go to work and detect the phone number and street address, if it's in there.

And why when I say "create a new appointment from 9 to 11 AM on Friday titled john smith visit at john smith's house" can it not look up John Smith's address from his contact in my phone?

Yeah, this is my very myopic complaint, because I do all of these things about 4 times per day as I get emails from new clients, add them to my contact list, and create an appointment with them. But Apple has like 250,000 employees... can't one of them just spend 3 days making these little fixes?

 

[Reverend Benny]

OK I've waited several years in hopes of some improvement. I don't feel like it's much to ask.

Apple knows if a phone number is an iPhone (because of iMessage compatibility) - why, when I add a contact directly from an email, does it assume the phone number is a home number instead of querying it and labeling it mobile if it's another iPhone number? And why when I add a contact in that way, do I FIRST have to add the name and email address, and only THEN do the data detectors go to work and detect the phone number and street address, if it's in there.

And why when I say "create a new appointment from 9 to 11 AM on Friday titled john smith visit at john smith's house" can it not look up John Smith's address from his contact in my phone?

Yeah, this is my very myopic complaint, because I do all of these things about 4 times per day as I get emails from new clients, add them to my contact list, and create an appointment with them. But Apple has like 250,000 employees... can't one of them just spend 3 days making these little fixes?

You can turn imessage off if you want, so not sure they always would know its an iPhone that's communicating.

 

[cameronjpu]

You can turn imessage off if you want, so not sure they always would know its an iPhone that's communicating.

Of course, and in those cases, it's OK. But what I want to know is, why doesn't apple try to help their users when the phone IS using iMessage to put it into the correct contact category.

 

[BigBlur]

why, when I add a contact directly from an email, does it assume the phone number is a home number instead of querying it and labeling it mobile if it's another iPhone number?

Yours defaults to 'home'? Mine defaults to 'mobile' whether it's actually a mobile number or not. I used to think it was odd that it wasn't smart enough to pick the 'iPhone' label (because of the iMessage thing you mentioned). That's what I pick if I know they have an iPhone, but at the end of the day, is there really a benefit to labeling these "correctly"? Isn't it just a personal preference on how you label them?

 

[cameronjpu]

Here's another fun example from today - I told siri to text a contact - "hey siri text john smith that i'm on the way and will be there i 15 minutes". Text failed. Why? Because the phone sent the text to the "home" phone number (which is a landline) instead of the mobile, which is an iPhone. Yes, I know that if I said "send a text to john smith mobile saying..." it would have worked. But why isn't Siri smart enough to assume, IF it's going to make an assumption and not ask, that it should send a text to a mobile number? Again... it knows that the mobile is an iPhone. It shows up in blue. Why would it text a home number without asking it it should do the mobile, or just do the mobile without asking?

 

[cameronjpu]

Yours defaults to 'home'? Mine defaults to 'mobile' whether it's actually a mobile number or not. I used to think it was odd that it wasn't smart enough to pick the 'iPhone' label (because of the iMessage thing you mentioned). That's what I pick if I know they have an iPhone, but at the end of the day, is there really a benefit to labeling these "correctly"? Isn't it just a personal preference on how you label them?

View attachment 2488782

Well yes, not everyone's world relies on having accurate information in the phone. But it costs me time if I'm trying to track someone down and I don't know which phone is a home and which is a mobile. And I also use my database to send marketing texts to my clients. So if I didn't enter the information correctly from the start, I would have a mess of a database right now. And frankly, after 20 years of making changes manually, I have let it slip a little recently as I do more with my voice, so it feels like more of an imposition to pick up the phone and make corrections. So now my database is getting less trustable, and that will certainly cost me money.

And yes, mine defaults to home. And even in my world, where I am dealing with older clients, it's still vastly more likely that I get contacted by cell, and people include their cell in their email signatures instead of home, generally. Yet, the iPhone still plays dumb and puts everything in home by default.

 

[TgeekB]

This is simply a case of if they do it one way people will complain and if they do it another way people will complain. Honestly, it doesn’t sound like a big lift.

 

[BenGoren]

This is simply a case of if they do it one way people will complain and if they do it another way people will complain. Honestly, it doesn’t sound like a big lift.

No, it really is a case of a profound lack of intelligence.

When Apple first developed these technologies, nobody expected actual intelligence. The logic they implemented was pretty decent for the time, clearly defined in code, and defensible — if nothing else, as you note, as a case of not being able to satisfy everybody so doing their best to navigate the least worst options.

But … now everybody expects (whether justified or not) computers to be at least as intelligent as, say, a grade school kid. Who can pass the bar exam with flying colors. ChatGPT, in other words.

So when our computers so spectacularly and idiotically do things that make no sense whatsoever, we get frustrated.

cameronjpu is spot on. Yes, when this code was written, it would have taken too much computational (and programmer) resources to do the sorts of things one would expect. But today, with “Apple Intelligence” powering the devices, and the compute cycles the phones themselves have?

It’s inexcusable.

Even back in the day, it would have been trivial to (for example) prioritize a mobile number for a text message over a “home” number. Today, it’s bloody obvious and should be the default. And, if you go out of your way to send a text message to a home number and have it fail, the iPhone should prompt you to retry with a mobile number. And it should (internally, privately) log which numbers you use in what circumstances (time of day, location, AirPods connected or not, etc.) to better pick a default choice — including knowing when there’s not enough confidence to make a choice and so to prompt you instead of making one for you.

It’s that sort of thing that I, at least, was expecting from Apple Intelligence. Not laughably bad summaries of my email and autocomplete that slows down my typing.

b&

 

[TgeekB]

No, it really is a case of a profound lack of intelligence.

When Apple first developed these technologies, nobody expected actual intelligence. The logic they implemented was pretty decent for the time, clearly defined in code, and defensible — if nothing else, as you note, as a case of not being able to satisfy everybody so doing their best to navigate the least worst options.

But … now everybody expects (whether justified or not) computers to be at least as intelligent as, say, a grade school kid. Who can pass the bar exam with flying colors. ChatGPT, in other words.

So when our computers so spectacularly and idiotically do things that make no sense whatsoever, we get frustrated.

cameronjpu is spot on. Yes, when this code was written, it would have taken too much computational (and programmer) resources to do the sorts of things one would expect. But today, with “Apple Intelligence” powering the devices, and the compute cycles the phones themselves have?

It’s inexcusable.

Even back in the day, it would have been trivial to (for example) prioritize a mobile number for a text message over a “home” number. Today, it’s bloody obvious and should be the default. And, if you go out of your way to send a text message to a home number and have it fail, the iPhone should prompt you to retry with a mobile number. And it should (internally, privately) log which numbers you use in what circumstances (time of day, location, AirPods connected or not, etc.) to better pick a default choice — including knowing when there’s not enough confidence to make a choice and so to prompt you instead of making one for you.

It’s that sort of thing that I, at least, was expecting from Apple Intelligence. Not laughably bad summaries of my email and autocomplete that slows down my typing.

b&

I disagree. The wild accusations you make are unfounded as technology and mobile devices have advanced at a rapid pace. Small details that require minute adjustments by humans are somehow “inexcusable”? 🤣 Do you have the background and knowledge to divulge just how easy it would be if you were in charge? Our society has become so dependent on technology that people literally cannot cope with having to do anything by themselves or accept what it takes to build an operating system that meets every demand.

 

[splifingate]

a grade school kid. Who can pass the bar exam

I have never, personally, met such an individual

 

[splifingate]

can't one of them just spend 3 days making these little fixes?

I'm sure they spend more time than that on such details!

That you and I (and a plethora of 'others') come to the Table with bespoke entry-points that it's really not possible to cater to The Crowd.

I spend a lot of time further-editing the Details, but that's just where I'm at, and where I feel comfortable....

 

[BenGoren]

I disagree. The wild accusations you make are unfounded as technology and mobile devices have advanced at a rapid pace. Small details that require minute adjustments by humans are somehow “inexcusable”? 🤣 Do you have the background and knowledge to divulge just how easy it would be if you were in charge?

Er … yes. Yes, I do.

And so does anybody who’s capable of writing several dozen lines of code without assistance.

Here, for example, is cameronjpu’s first complaint:

"Apple knows if a phone number is an iPhone (because of iMessage compatibility) - why, when I add a contact directly from an email, does it assume the phone number is a home number instead of querying it and labeling it mobile if it's another iPhone number?”

Suppose you’re the programmer tasked with this. You already have the existing code to add a contact directly from an email. That code has already parsed out the phone number. Today, it just blindly adds that number to the contact with the “Home” tag. Your job is to determine whether or not it might better deserve the “iPhone” tag.

The old code (simplified into pseudocode) looks something like this:

with NewContact.Phone
.Number = parsedNumber
.Tag = “Home"
end with

Your new code will look like:

with NewContact.Phone
.Number = parsedNumber
if iMessage.Verify(parsedNumber) then
.Tag = “iPhone"
else
.Tag = “Home"
end if
end with

There — not so bad, was that?

Note that the “iMessage.Verify()” function would be the exact same one that already exists and is used by iMessage to determine if this is a blue-bubble or green-bubble conversation.

And, of course, this is all pseudocode. The language determines the syntax. The API determines the function names and parameters. Real code is going to include error handling and other sorts of “housekeeping” — and there will be much more such code than “actual doing of things” code. And you’re going to spend most of time studying the existing code, researching your options, testing and validating the results, and (especially) documenting everything you did and why. Far more time, in fact, than you will on actually writing code.

But, since coding the “actual doing of things” in this instance that you'll spend maybe 5-10% of time on really is as trivial as what I have above, you can tell that this particular job is an especially easy one — indeed, one that would be very well suited for a first “real” task for a newly-hired engineer on the team.

The other items on cameronjpu’s wish list are all of comparable difficulty to implement. The existing code is already doing all the “heavy lifting,” and you really don’t need more than this sort of one-banana code for the final polishing.

Cheers,

b&

 

[cameronjpu]

This is simply a case of if they do it one way people will complain and if they do it another way people will complain. Honestly, it doesn’t sound like a big lift.

I'm not following. Why would people complain if the phone used its already existing ability to detect an iPhone to file it as a mobile? Or are you talking about something else?

 

[cameronjpu]

No, it really is a case of a profound lack of intelligence.

...

It’s that sort of thing that I, at least, was expecting from Apple Intelligence. Not laughably bad summaries of my email and autocomplete that slows down my typing.

b&

Ever wonder why, when you're dictating a message, a random word in the middle of the sentence is capitalized? Or when you say the sentence "I didn't think I'd ever hear something like that" it uses "Heare" because that's the last name of a contact in your contact list? Or you dictate and the result is "I didn't realize you're daughter was already 8"

Pretty smart. All I ask in those situations is a level of grammar checking equivalent to Word in 2007. Why is that so hard?

 

[TgeekB]

And, of course, this is all pseudocode. The language determines the syntax. The API determines the function names and parameters. Real code is going to include error handling and other sorts of “housekeeping” — and there will be much more such code than “actual doing of things” code. And you’re going to spend most of time studying the existing code, researching your options, testing and validating the results, and (especially) documenting everything you did and why. Far more time, in fact, than you will on actually writing code.

Cheers,

b&

At least you’re honest. Pseudocode. You have no idea what the actual coding is or how that coding affects other functions within the system. You’re just guessing.
I’m sure it’s doable. I’m sure the programmers at Apple have the knowledge to accomplish it. What we don’t know is their task list, what they are working on, how long it is taking them and where such a lowball change fits in the whole. Very low is my guess. So “inexcusable” has no meaningful connection, except in someone’s head on the internet, to what actually matters or occurs. Programmed are not sitting around reading these forums looking for something to do.

 

[TgeekB]

I'm not following. Why would people complain if the phone used its already existing ability to detect an iPhone to file it as a mobile? Or are you talking about something else?

Read my response above.

 

[BenGoren]

At least you’re honest. Pseudocode. You have no idea what the actual coding is or how that coding affects other functions within the system. You’re just guessing.

Squeeze me?

What actual developer would even think to use anything other than pseudocode to describe the logic involved? I mean, you don’t think that even an Apple developer chiming in on this thread would start with a git pull, do you?

(And, no, “a git pull” has nothing to do with tying ropes to unpleasant people. Usually … exceptions have been known to have been made … )

Validating a phone number isn’t rocket surgery. That’s the point we’re all trying to explain to you.

The particular language used for the implementation is completely irrelevant to this discussion. Just … so beyond irrelevant that I’m not sure how to address your concerns that it’s some showstopping impediment. Phone number validation happens all over the place in Apple products … there’s no more worry about “how that coding affects other function within the system” than there is for adding spellcheck to a text input.

I don’t know why what we have is failure to communicate.

This is some very basic, very elementary fit-and-finish polishing work that Apple used to excel at — and, in so many ways, still does.

That’s the frustration: how can they get so many things right, yet miss so many of these very easy things?

It’s like buying some nice high-end sedan, but the volume knob is chrome-plated plastic that’s already scratched by the time you take it for a test drive. I mean, sure, the car drives great and the stereo sounds awesome and the knob looks pretty in pictures of the dash … but, really?

b&

 

[TgeekB]

Squeeze me?

What actual developer would even think to use anything other than pseudocode to describe the logic involved? I mean, you don’t think that even an Apple developer chiming in on this thread would start with a git pull, do you?

(And, no, “a git pull” has nothing to do with tying ropes to unpleasant people. Usually … exceptions have been known to have been made … )

Validating a phone number isn’t rocket surgery. That’s the point we’re all trying to explain to you.

The particular language used for the implementation is completely irrelevant to this discussion. Just … so beyond irrelevant that I’m not sure how to address your concerns that it’s some showstopping impediment. Phone number validation happens all over the place in Apple products … there’s no more worry about “how that coding affects other function within the system” than there is for adding spellcheck to a text input.

I don’t know why what we have is failure to communicate.

This is some very basic, very elementary fit-and-finish polishing work that Apple used to excel at — and, in so many ways, still does.

That’s the frustration: how can they get so many things right, yet miss so many of these very easy things?

It’s like buying some nice high-end sedan, but the volume knob is chrome-plated plastic that’s already scratched by the time you take it for a test drive. I mean, sure, the car drives great and the stereo sounds awesome and the knob looks pretty in pictures of the dash … but, really?

b&

Fit and finish polishing work gets done at the end when there’s time. I don’t understand your frustration. It’s not hurting Apple or its users in any way. Again, you have zero idea what their workload is. There are so many aspects and teams that go into building an iPhone and the AI project which is top priority. We can all nitpick and say we would do it different but we/you aren’t the experts. Our society is so bent in having everything instantly and this conversation is a prime example. It will come when it comes and, in the meantime, a child can still get by without it.

 

[BenGoren]

Fit and finish polishing work gets done at the end when there’s time.

Eh … no.

Make, that, “Hell no!”

Fit and finish is considered from the get-go, and a non-trivial part of what you spend your time on. Slapdash “well it’s good enough for now; we’ll fix it later” in any discipline is something done only by amateurs, by unprofessional lowballers, or in extreme emergency triage situations.

Why?

Because, if you don’t have the time or resources to do it right the first time, you sure as anything don’t have the time or resources to fix it later.

A general contractor installing kitchen cabinets once made the point rather well. I remarked as to how he always appeared to be moving at an especially relaxed pace — almost lazy. But every time I turned around, he had finished some huge chunk of the project. He explained that his father taught him, “Slow is smooth. Smooth is fast. If you want to work fast, work slow.” You don’t make anywhere near as many mistrakes that need fixing, and the solid foundation you lay at the early stages makes everything else you do afterwards that much easier.

I don’t understand your frustration. It’s not hurting Apple or its users in any way.

Except that cameronjpu has repeatedly explained how much of his time these rough edges cost him personally.

Maybe you yourself never go near these rough edges, or you only do so rarely enough that they’re inconsequential to you. Or maybe you like the pain.

But he(?)’s running into them all the time. And that hurts.

Again, you have zero idea what their workload is.

Apple is the most valuable company in the history of civilization. Countless other companies in other industries manage to polish equivalent rough edges — including Apple itself in the past.

Not paying attention to these details is a choice on Apple’s part. Whatever they’re prioritizing these days, it isn’t this sort of quality control.

There are so many aspects and teams that go into building an iPhone and the AI project which is top priority.

AI is the current top leadership’s top priority — and, yet, all we get are non-stop announcements of great new AI features coming soon, followed a week later about how the previous round of great new features we were supposed to have next month are now pushed back by a year or two.

Apple is struggling right now. They’re still the best choice overall, especially for long-time users … but they’re running out of momentum.

It’s happened before, to pretty much every big player in the industry, at one time or another — Apple included. Apple has recovered (and how!). Others haven’t (DEC, Compaq, Commodore), and others have transformed themselves into something just as big but essentially unrecognizable from before (IBM, Dell).

Time will tell what Apple’s future holds.

We can all nitpick and say we would do it different but we/you aren’t the experts.

Pro tip: you don’t need to be an expert to tell when a product’s rough edges is giving you a rash.

Nor are you under any burden to shut up, suck it up, and suffer silently when a product you otherwise like keeps giving you a rash.

Nor does a wise expert want you to — or even make excuses for why the rough edge is necessary.

Heck, that’s what powered Apple’s rise to greatness in the first place: making stuff that “just works,” and getting rid of all the rough edges.

Our society is so bent in having everything instantly and this conversation is a prime example. It will come when it comes and, in the meantime, a child can still get by without it.

You know? That’s exactly what the critics of the Mac said in 1984. Typing commands in DOS just takes a moment. Why do you need a “gooey” with a mouse pooping on your dinner menu? It’s all just instant gratification — besides which, it’s so much faster to type “RUN D:\CODE\MS\DOC.EXE /EDIT /SET-LINE=1437 /DELETE-LINES=1 /INSERT=‘concerns about product @ITONquality@ITOFF from the previous meeting. I look forw-‘ /AUTO-PRINT=TRUE” than to have to click and scroll and look for stuff. Just ask your kid if you’re too old to figure it out yourself.

b&

 

[TgeekB]

Eh … no.

Make, that, “Hell no!”

Fit and finish is considered from the get-go, and a non-trivial part of what you spend your time on. Slapdash “well it’s good enough for now; we’ll fix it later” in any discipline is something done only by amateurs, by unprofessional lowballers, or in extreme emergency triage situations.

Why?

Because, if you don’t have the time or resources to do it right the first time, you sure as anything don’t have the time or resources to fix it later.

A general contractor installing kitchen cabinets once made the point rather well. I remarked as to how he always appeared to be moving at an especially relaxed pace — almost lazy. But every time I turned around, he had finished some huge chunk of the project. He explained that his father taught him, “Slow is smooth. Smooth is fast. If you want to work fast, work slow.” You don’t make anywhere near as many mistrakes that need fixing, and the solid foundation you lay at the early stages makes everything else you do afterwards that much easier.



Except that cameronjpu has repeatedly explained how much of his time these rough edges cost him personally.

Maybe you yourself never go near these rough edges, or you only do so rarely enough that they’re inconsequential to you. Or maybe you like the pain.

But he(?)’s running into them all the time. And that hurts.



Apple is the most valuable company in the history of civilization. Countless other companies in other industries manage to polish equivalent rough edges — including Apple itself in the past.

Not paying attention to these details is a choice on Apple’s part. Whatever they’re prioritizing these days, it isn’t this sort of quality control.



AI is the current top leadership’s top priority — and, yet, all we get are non-stop announcements of great new AI features coming soon, followed a week later about how the previous round of great new features we were supposed to have next month are now pushed back by a year or two.

Apple is struggling right now. They’re still the best choice overall, especially for long-time users … but they’re running out of momentum.

It’s happened before, to pretty much every big player in the industry, at one time or another — Apple included. Apple has recovered (and how!). Others haven’t (DEC, Compaq, Commodore), and others have transformed themselves into something just as big but essentially unrecognizable from before (IBM, Dell).

Time will tell what Apple’s future holds.



Pro tip: you don’t need to be an expert to tell when a product’s rough edges is giving you a rash.

Nor are you under any burden to shut up, suck it up, and suffer silently when a product you otherwise like keeps giving you a rash.

Nor does a wise expert want you to — or even make excuses for why the rough edge is necessary.

Heck, that’s what powered Apple’s rise to greatness in the first place: making stuff that “just works,” and getting rid of all the rough edges.



You know? That’s exactly what the critics of the Mac said in 1984. Typing commands in DOS just takes a moment. Why do you need a “gooey” with a mouse pooping on your dinner menu? It’s all just instant gratification — besides which, it’s so much faster to type “RUN D:\CODE\MS\DOC.EXE /EDIT /SET-LINE=1437 /DELETE-LINES=1 /INSERT=‘concerns about product @ITONquality@ITOFF from the previous meeting. I look forw-‘ /AUTO-PRINT=TRUE” than to have to click and scroll and look for stuff. Just ask your kid if you’re too old to figure it out yourself.

b&

Ignoring your wall of text.
End result it switching home to mobile takes, let’s say, 15 seconds. He states this happens 4 times per day. 15 seconds times 4 equals….1 minute. 1 minute! If this took 1 hour I would have more sympathy and totally agree.
We will have to agree to disagree. Enjoyed the convo. All the best!

 

[BenGoren]

Ignoring your wall of text.
End result it switching home to mobile takes, let’s say, 15 seconds. He states this happens 4 times per day. 15 seconds times 4 equals….1 minute. 1 minute! If this took 1 hour I would have more sympathy and totally agree.
We will have to agree to disagree. Enjoyed the convo. All the best!

Ah, yes.

What’s all the fuss about? It only takes a moment to remove the fly, and you don’t even lose a full spoon’s worth of your soup.

Some — but, clearly, not all — of us place importance on professionalism and pride of workmanship.

b&

 

[cameronjpu]

Read my response above.

Your response above appears to just say that programming is hard, as an excuse for why any feature would not be added to the phone. What I’m saying is, somebody is seriously prioritizing what things should shouldn’t be done when small issues such as these are left, ignored, for years. Honestly, you can save us your words because we know what you’re going to say in response to everything. Your answer is “there’s no way for us to know why Apple isn’t doing these things but it’s probably because they are busy doing other things“. Thank you.