Apple Preparing to Add ChatGPT Integration to Siri

[MacFan23]

I wonder if widespread availability of AI via the Apple ecosystem will begin to let the air out the AI bubble. People will be to use it and see that it is really just a parlor trick.

I always find these comments very confusing. I use AI everyday in my life and I find new ways to utilise it the more I use it. It's made genuine improvements to my life and workflow.

It's absurd to me that comments like this are so prevalent and popular.

 

[WarmWinterHat]

I always find these comments very confusing. I use AI everyday in my life and I find new ways to utilise it the more I use it. It's made genuine improvements to my life and workflow.

It's absurd to me that comments like this are so prevalent and popular.

I generally see tech workers say things like that about AI, but it's not useful to everyone and not applicable to every situation. My company, for example, has a blanket ban on using AI for anything whatsoever, and even has the sites blocked network-wide because none of them are 100% HIPAA compliant. I don't work in tech, or anything tech-adjacent.

I've personally tried chatGPT, and others, many times, and haven't been able to find any use for them. I get bored after a few times and move on.

 

[Northgrove]

Before you start complaining: NOBODY is forcing y’all to use this feature.

True, and what's more - providing support for a cloud service is also having minimal impact on resource use or iOS footprint since the hard work is done remotely. This is "just" integrating with API endpoints. The real, heavy load is on OpenAI's side on how to deal with millions of new AI requests... We're lucking if Siri is getting the 4o model. I'm thinking 4o-mini, or a custom finetune for iOS and even smaller.

 

[The Game 161]

seeing people talk about opting out is laughable. seriously get with the times and move on. tech is changing all the time and none of this you have to use but dismissing it just seems odd.

 

[deeside81]

So when is it coming out?

 

[Wando64]

Ask Siri something simple. "I'm sorry, I can't do that."
Ask ChatGPT to solve a textual math problem: "Sure! Using this formula and those calculations, this comes to...", and remembers you suck at math and have a cat.

Example: Ask Siri: What's 1958 $58 in 2024? Siri: I found this on the web! And shows a link to some sports ball team?
Ask ChatGPT the same question: $58 in 1958 is approximately equivalent to $614.10 in 2024, considering average inflation rates over that period.

Ask Siri: What do you call the kinds of books published by Paladin Press? *shows link to Scholastic dot com*
ChatGPT: Books published by Paladin Press are often referred to as "how-to" manuals on controversial and often extreme subjects. These include topics such as self-defense, survivalism, weaponry, guerrilla warfare, and other unconventional or underground activities. The term "anarchist literature" or "military and survival manuals" are also commonly used to describe the genre of books that Paladin Press specialized in.

This is all very true, as long as you ignore the numerous instances of GPT making up the answer.
I have discovered this on numerous occasions when I was double checking some results.
In those cases asking GPT “are you sure?” will elicit an apology followed by a different answer (also made up).
Keep on asking “are you sure?” or “I think you are wrong” and you will get more and more apologies followed by more and more made up answers.
It would be comical if it wasn’t that many people rely on these answers.

 

[WarmWinterHat]

seeing people talk about opting out is laughable. seriously get with the times and move on. tech is changing all the time and none of this you have to use but dismissing it just seems odd.

I don't personally care if I seem old; I see no point in having a conversation with my phone when it's not another person. I don't use Siri or any other voice assistant, and have never found them useful. Some "AI" features might be useful, like photo touchups, suggestions, summaries, but not conversational/generative AI, for me.

...that and the lack of ethics of OpenAI; scraping anything and everything without permission of the site owner or creator of content. That is some garbage that I can't look past.

 

[jonnysods]

I use ChatGPT in a controlled manner, as in its own separate app. I’d like Siri to be smarter and give better results, but I want to see how it behaves on device and what data leaves my phone. This all feels a little rushed.

 

[m4mario]

Make no mistake. Apple is working on its own ChatGPT. Where its struggling is acquiring the globules of data required while avoiding the legal and PR consequences of just scraping every possible data in sight without acknowledgement from the content owners.

Other companies have decided to get market share today and deal with the consequences later.

 

[cgs1xx]

I tend to switch Siri off as soon as I buy a new iPhone, I've never needed it for anything 🤷🏻

 

[Kar98]

I generally see tech workers say things like that about AI, but it's not useful to everyone and not applicable to every situation. My company, for example, has a blanket ban on using AI for anything whatsoever, and even has the sites blocked network-wide because none of them are 100% HIPAA compliant. I don't work in tech, or anything tech-adjacent.

I work in I.T. and need to follow HIPAA standards, too, and this company also has banned AI use company-wide. Mostly because they realize there's entire strata of employees (85% of management) that is completely useless and could be replaced with $20/month chat scripts.

 

[nt5672]

Came here to see all the luddite curmudgeons complaining about "parlor trick" AI that "nobody asked for" and that they "better be able to opt out of". I was not disappointed.

If you people were the majority we'd be using Apple typewriters right now instead of Apple computers.

And if you cannot understand the difference between machines that do what we tell them i.e. computers replacing typewriters, and LLMs that just regurgitate words that may or may not be true, but that are definitely biased based on who trained them, then the world will suffer, but not because of LLMs, but because of people.

 

[Kar98]

This is all very true, as long as you ignore the numerous instances of GPT making up the answer.
I have discovered this on numerous occasions when I was double checking some results.
In those cases asking GPT “are you sure?” will elicit an apology followed by a different answer (also made up).
Keep on asking “are you sure?” or “I think you are wrong” and you will get more and more apologies followed by more and more made up answers.
It would be comical if it wasn’t that many people rely on these answers.

Nobody said to rely on it? Even ChatGPT says: "ChatGPT can make mistakes. Check important info."
Also I think you're asking it wrong

 

[iphonehype]

I generally see tech workers say things like that about AI, but it's not useful to everyone and not applicable to every situation. My company, for example, has a blanket ban on using AI for anything whatsoever, and even has the sites blocked network-wide because none of them are 100% HIPAA compliant. I don't work in tech, or anything tech-adjacent.

I've personally tried chatGPT, and others, many times, and haven't been able to find any use for them. I get bored after a few times and move on.

Microsoft Copilot îs HIPAA complaint from August this year, also companies just banning AI will fall behind the AI race, very fast. Also, I hardly use Google anymore to search, I mostly use ChatGPT.

I use Google to understand something e.g. What is X and how does it work with Y?

With ChatGPT I can upload a picture and say tell me more, or just have a voice conversation. Download the ChatGPT app and give it another go the tech has come far since you may have last used it

 

[iphonehype]

Make no mistake. Apple is working on its own ChatGPT. Where its struggling is acquiring the globules of data required while avoiding the legal and PR consequences of just scraping every possible data in sight without acknowledgement from the content owners.

Other companies have decided to get market share today and deal with the consequences later.

Did Tim phone you?

 

[Kar98]

I use Google to understand something e.g. What is X and how does it work with Y?

Except now instead of giving useful results, the top result is some hallucinated AI garbage followed by TEMU ads...

So might was well skip the middle man and ask ChatGPT directly.

 

[WarmWinterHat]

Microsoft Copilot îs HIPAA complaint from August this year, also companies just banning AI will fall behind the AI race, very fast. Also, I hardly use Google anymore to search, I mostly use ChatGPT.

I use Google to understand something e.g. What is X and how does it work with Y?

With ChatGPT I can upload a picture and say tell me more, or just have a voice conversation. Download the ChatGPT app and give it another go the tech has come far since you may have last used it

You missed a couple points in my post. 1) I don't want to have a conversation an AI. Give me the data, without the superfluous fake chatter, and 2) openAI is unethical as hell, and I refuse to touch their service.

I don't use Google either, and have moved to DuckDuckGo, but I don't use search at all much anymore and just visit a handful of sites.

 

[PsykX]

Before you start complaining: NOBODY is forcing y’all to use this feature.

There's a lot of myths to debunk regarding Apple Intelligence.

Here's what most people think (wrongfully) :
1. You are forced to use it.
2. Apple Intelligence is in fact just a rebranded version of ChatGPT.
3. Apple Intelligence sends everything to servers.
4. Apple Intelligence is already available if you upgrade your phone (thanks to the big "Hello, Apple Intelligence" right at the top of Apple's iPhone 16 Pro page)

 

[nt5672]

I always find these comments very confusing. I use AI everyday in my life and I find new ways to utilise it the more I use it. It's made genuine improvements to my life and workflow.

It's absurd to me that comments like this are so prevalent and popular.

Using LLMs as a better search engine, showing sources, getting ideas is one thing. Using them to do actual work like producing code, text, images, etc. is just plain stupid. This will only lead to the person being replaced. There will be no need for the person in the middle to turn out mediocre work.

 

[dannys1]

I generally see tech workers say things like that about AI, but it's not useful to everyone and not applicable to every situation. My company, for example, has a blanket ban on using AI for anything whatsoever, and even has the sites blocked network-wide because none of them are 100% HIPAA compliant. I don't work in tech, or anything tech-adjacent.

I've personally tried chatGPT, and others, many times, and haven't been able to find any use for them. I get bored after a few times and move on.

I mean it's the first time i've ever seen my technophobe girlfriend use a new technology regularily. If you can't find use for a tool which you can talk to in natural language and will talk back and knows almost everything humans have ever written, can to a degree think and calculate, code and devise then i'm not sure what to tell you.

Like the internet I struggle to imagine my life without it now.

 

[LIVEFRMNYC]

All this A.I greatness for Siri, and I will still mostly use Siri for "Set Timer" "Call Mom" and "Set Alarm.

 

[citysnaps]

People getting a case of the shakes about AI reminds me of when the WorldwideWeb was introduced long ago.

 

[WarmWinterHat]

I mean it's the first time i've ever seen my technophobe girlfriend use a new technology regularily. If you can't find use for a tool which you can talk to in natural language and will talk back and knows almost everything humans have ever written, can to a degree think and calculate, code and devise then i'm not sure what to tell you.

Like the internet I struggle to imagine my life without it now.

I guess I just don't get it, then. I can also not use the internet for several days, or weeks, when out hiking/backpacking and it's not a big deal, so imagining my life without either isn't hard. I'd rather search exactly what I'm looking for, and it be given back to me in a listed format without all the human-esque superfluous chatter.

People getting a case of the shakes about AI reminds me of when the WorldwideWeb was introduced long ago.

No shakes here. I just think it's boring and not interesting enough to care about.

I do think the ethics of AI companies scraping data without an okay is a real problem, though. No one, anywhere, should have their data used without explicit opt-in permission.

 

[ThailandToo]

True, and what's more - providing support for a cloud service is also having minimal impact on resource use or iOS footprint since the hard work is done remotely. This is "just" integrating with API endpoints. The real, heavy load is on OpenAI's side on how to deal with millions of new AI requests... We're lucking if Siri is getting the 4o model. I'm thinking 4o-mini, or a custom finetune for iOS and even smaller.

Then why can’t iPhones run it unless you have 15 Pro or newer? Even people who bought an iPhone 15 a few months ago have an iPhone that will not run ChatGPT = “Apple non-intelligence” because Apple wants to sell more iPhones. That is the real joke in all of this.

 

[Kar98]

I guess I just don't get it, then. I can also not use the internet for several days, or weeks, when out hiking/backpacking and it's not a big deal, so imagining my life without either isn't hard.

I'd rather search exactly what I'm looking for, and it be given back to me in a listed format without all the human-esque superfluous chatter.

You can request ChatGPT to behave just like that, if you want. The whole hey buddy routine is just fluff and decoration. If you want just a list of results, you can get that. If you want a uWu senpai style convo, you can have that. It's just like zombo dot com!