Apple Delays Major iOS 19 Feature Announcement

[brijazz]

Better later and better than the other companies

I used to feel this was "the Apple way". But I increasingly feel it's become "build something and let it languish." Siri COULD have been improved over the years - even before AI become a thing that other phones have - but they didn't bother.

 

[TechnoTiger3000]

It’s interesting that for some folks Apple Intelligence can’t come soon enough, yet for others they are eager just to disable it immediately after release.

Folks want things to work and apple only focuses on becoming a fantasy marketing machine.

 

[HouseLannister]

Gurman was wrong about 18.4, so now he’s hedging his bets around 19.4. It’s okay to say things are in development and may not come out and you don’t know for sure. Nothing has been delayed based on Apple’s announcements, only Gurman’s.

 

[TechnoTiger3000]

Let’s face it: Apple’s a hard company more than they are a software company. One could argue the best new thing they introduced since Tim Cook took over as CEO was Apple Silicon. When it comes to AI nobody is looking to Apple to lead.

*The only thing that seems to function is the hardware division.

They are still a software company, but that devision seems solely focused on bs niche projects (you can move your mac mouse to your iPad, jeej?) and making it as hard as possible for customers to use better software… pathetic.

They can’t even get their spellcheck close to useful functioning in 2025, wish them good luck with APpLe InTelLigENcE.

 

[nfl46]

Lol. It will go from 19.4 to 20.4 to 21.4.

 

[Cloudyskies22]

You sound like my grandfather who told me we’d all be flying around in flying cars by now because he saw the film Metropolis and had a subscription to Popular Science. You cornucopian utopianists always amuse me. AI evangelists are simply that….evangelists. And because of that, by default, are cultists.

Ever see what the discovery of and harnessing of electricity did for civilization?


AI will be on that level and there is nothing that compares to that. I know it’s easy for all of you who have zero vision or understanding to dismiss everything that comes along but AI is a very vastly different beast to anything aside from electricity. You’ll see within 2 years how it will quickly become profound. This year agents come, they’ll be AIs working to do specific things using any computer interface like humans do. In a few years they’ll be using robots not just digital interfaces. But the real thing is putting millions of PhDs in every field known, all working 24/7 on a task in the cloud. Enjoy

 

[Cloudyskies22]

Amazon always had a “product” to showcase for all the VC and Wall Street funding for all that time. And after every book sold , after every warehouse made, after every delivery truck purchased or leased, after every cloud contract signed to AWS, there was evidence that Amazon would finally hit “escape velocity” and become self sufficient. Not so with AI. Not with OpenAI, not with Anthropic, not with xAI, not with Perplexity. Not even Gemini. Yes, Google is losing money with Gemini. It’s only masked by Googles other money making schemes. For God’s sake….OpenAI loses money on their APIs much less their app or their website. Every prompt into their APIs and every output from their APIs per user loses OpenAI money. Amazon always had an infrastructure product to show Wall Street they were more than an online book store. Jeff Bezos, however you want to view him, and I view him VERY negatively, at least showed a vision of Amazon as more than an online “Barnes and Noble”. Sam Altman on the other hand is showing nothing to Wall Street but an AI hallucination to entice money from them long enough for him to eventually sell the company to a Microsoft or Oracle, etc, when the inevitable AI. bubble bursts.

AI will transform the world in the next 5 years. That’s a fact. By the end of this year the usage of LLMs will increase monumentally as capabilities continue to rapidly increase and as the agent systems come online.

 

[xyz01]

Better later and better than the other companies

It's later than the competition, and as far as features and quality goes it's not going to be near the competition.

Where it has the advantage is privacy.

Of course, coming from Norway where we still have the keyboard from the original release(*) it's probably going to be nonexistent. Just like for keyboards we use MS and Google keyboards, for AI we'll probably have to use their products as well - so the main feature for me is being able to use another AI.

(*) No swiping here with Apple, so I'm thankful for MS SwiftKey not ceasing to exist - and Google having another alternative.

 

[hans1972]

It honestly makes Siri seem like it’s the exact same thing it has been since it launched in 2011 which is basically what Siri is.

Siri isn't using Apple Intelligence and isn't supported by a LLM. It is still essential the same technology as when released.

Some people seems to think Siri has changed it core technology.

 

[hans1972]

Yeah I honestly think if OpenAI starts releasing hardware it could get real bad for Apple real fast. OpenAI will have AGI-level AI by late 2027

If you believe this, there isn't anything Apple can do which will stop OpenAI from dominating. You should prepare to move away from the Apple ecosystem by 2 years.

 

[hans1972]

Also the energy costs of AI are of ZERO consequence. We need to use ALL the energy necessary for AI because AI will be the only thing that can stop climate change. Humans cannot do it alone in time. AGI and ASI will build out fusion plants astronomically faster. They’ll solve innumerable climate issues clear across the board.

These are things that will be happening. Oh, by the way, AI will cure all human diseases within 15 years. Enjoy. And by the way, the very reason you know AI is the real deal is specifically because the cheaters in finance and on Wall Street are pouring literally hundreds of billions into it. Including the US government. Guess what buddy—these people do their due diligence.

You should apply for the position as Comical Ali at OpenAI.

 

[hans1972]

Go search how much Amazon was losing per year before they made a single cent in profit.

Amazon lost literally billions per year before becoming profitable. It’s called infrastructure build out.

Amazon lost a little over $3 billion in total from 1994-2002. Only on year did they loose more than $1 billion, in 2000.

 

[Cloudyskies22]

If you believe this, there isn't anything Apple can do which will stop OpenAI from dominating. You should prepare to move away from the Apple ecosystem by 2 years.

If OpenAI doesn’t make a phone which I don’t really think they will, Apple is safe for a very long time selling phones if nothing else.

Phones aren’t going anywhere anytime soon because they have cellular and increasingly satellite communication and they have a great screen. Even if smart glasses come and have translucent screens, they still won’t replace a smartphone because translucent screens are garbage for anything like media consumption.

 

[Cloudyskies22]

Siri isn't using Apple Intelligence and isn't supported by a LLM. It is still essential the same technology as when released.

Some people seems to think Siri has changed it core technology.

Buddy Siri is in fact using Apple intelligence now. The issue is that Apple intelligence isn’t intelligent, it’s dumb and worthless.

If your siri has a glowing ring around it, it is using Apple intelligence. They supposedly helped it better understand stumbled queries, and better more natural responses.

The issue is that it’s still embarrassingly god awful even with that. Meanwhile we look at all the major AI companies and every one of them has something that seems decades ahead of Siri lmao. By the way, when Alexa releases their new Alexa using Claude it will truly make Siri look so useless.

 

[smulji]

By then, who knows, 16GB RAM could easily be the bare minimum on iPhones

It should be that by now, if not now then starting with iPhone 17 series.

 

[Wx_Man]

If you believe this, there isn't anything Apple can do which will stop OpenAI from dominating. You should prepare to move away from the Apple ecosystem by 2 years.

To what ? Android ? LOL…then you’re moving to Samsung’s ecosystem because they are the only one that’s viable as a consumer company vis a vis Android. To Microsoft…LOL! They’ve already given up on phones, tablets, remember the Zune music player….right…nobody does. They’re killing off more Surface products. Good luck going there. The Humane AI pin is now dead. So….again….what is this mythical ecosystem you’re extolling people to go to in two years if not Apple ?

 

[smulji]

There is no killer app or hardware left to produce

I'm still bullish on AR glasses, smart home automation, and health tech

 

[Wx_Man]

To what ? Android ? LOL…then you’re moving to Samsung’s ecosystem because they are the only one that’s viable as a consumer company vis a vis Android. To Microsoft…LOL! They’ve already given up on phones, tablets, remember the Zune music player….right…nobody does. They’re killing off more Surface products. Good luck going there. The Humane AI pin is now dead. So….again….what is this mythical ecosystem you’re extolling people to go to in two years if not Apple ?

Oops…my bad. I see now your reply was to Cloudyskies silly take. My bad. I see where you were heading

 

[Wx_Man]

this is getting embarrassing. can't they just buy an AI company and crowbar it in a this point.

Who ? And at what valuation ? Every single AI company is WILDLY overvalued. Apple has already bought at least 2 different AI firms that were and are in the forefront at shrinking AI models down to fit in the NPU of every Apple Silicon based device with standard amounts of RAM on base systems. That’s an EXCEEDINGLY hard task to do as that ChatGPT 4.o or Anthropic Claude or Meta’s Llama AIs need scads of GPU RAM and have 100s of billions of parameters to be as “intelligent” as they are now. You’re asking Apple to do that on an iPhone with an NPU that maxes out at 18 TOPS with 8gb of shared RAM. If you think it’s that easy….then get to coding, because Apple needs your help…LOL.

 

[Fll01]

Disappointing. But I expected nothing else

 

[macmac30]

Bloomberg's Mark Gurman last year reported that Apple was planning a "more conversational" version of Siri for iOS 19.4, powered by "more advanced large language models." However, in his Power On newsletter today, he said that development of this more ChatGPT-like Siri is now running behind schedule within the company.

Gurman said Apple is no longer planning to unveil the more conversational Siri at WWDC 2025 in June this year, and it appears that the feature is no longer on track to launch as part of iOS 19.4 in March or April next year. Instead, he said that some Apple software engineers believe that the Siri overhaul will not launch until at least iOS 20.

This is yet another sign that Apple remains behind OpenAI in the generative AI space.

Gurman still expects iOS 19 to include some changes to Siri's underlying architecture that will allow it to handle more advanced queries.

In the meantime, iOS 18.2 added ChatGPT integration to Siri, and Google Gemini integration should follow in a later software update. And with iOS 18.5, which should be released in May, Siri is expected to gain on-screen awareness, understanding of personal context, and deeper per-app controls. Those features were initially expected to arrive with iOS 18.4, which further underscores Apple's internal delays in this area.

Article Link: Apple Delays Major iOS 19 Feature Announcement

I would be happy if Siri a) didn’t ask me who I am several times per week, b) listened and responded to my queries without me having to raise my voice and ask several times, and c) didn’t randomly pause music several times per hour on my newer gen large paired HomePods…

The bar is pretty low. Apple, are you up to it???

 

[Cloudyskies22]

Who ? And at what valuation ? Every single AI company is WILDLY overvalued. Apple has already bought at least 2 different AI firms that were and are in the forefront at shrinking AI models down to fit in the NPU of every Apple Silicon based device with standard amounts of RAM on base systems. That’s an EXCEEDINGLY hard task to do as that ChatGPT 4.o or Anthropic Claude or Meta’s Llama AIs need scads of GPU RAM and have 100s of billions of parameters to be as “intelligent” as they are now. You’re asking Apple to do that on an iPhone with an NPU that maxes out at 18 TOPS with 8gb of shared RAM. If you l think it’s that easy….then get to coding, because Apple needs your help…LOL.

Every single AI company is actually wildly undervalued. Compared to Apple, OpenAI is worth easily 500 times maybe 1,000 times what Apple is worth. That’s just in the next decade. In 15 years OpenAI will be worth 1 million times what Apple could ever hope to be.

AGI and ASI has essentially infinite value to humanity. It’s essentially the tree of life or the holy grail. Whatever mythical thing you want to attach it to, it’s that in value. It will replace all human jobs, it will cure all diseases, it will solve problems and make scientific discoveries and theories we couldn’t comprehend.

For example, Einstein made multiple theories that science could not empirical prove for about 100 years after he made them. Now imagine the theories AI will make that won’t be able to immediately be proven experimentally, but nevertheless, be true? It will be unimaginable. We will also trust they’re true because it will become clear that it’s highly accurate at theory accuracy.

It’s like if you could copy human intelligence millions and billions of times and have that shoved into a number of large data centers working 24/7 on certain tasks. The advancement will be unreal. The only things getting in the way will be physical experimentation and human society. We will slow it, but even still, it will be unbelievable advancement.

I get that it’s still very hard for a lot of people to accept or believe, but I’d say by this time next year people will finally start understanding what’s going to happen. Because this year a lot of AI agents will be starting to come online which will really get people to understand that this **** is real.

 

[rondo]

At this point they should just give up and end this misery.

 

[uacd]

more ChatGPT-like Siri

It must already be Grok-like or even better. I find GPT to often miss the context and fail to actually “think”. Also it is constantly unsure and often tells me “if you are not sure it is better…”.

With Grok it feels like I am interacting with Harvard professor and GPT is more like Master of Science.

Looking forward for REAL Apple Intelligence that understands contexts and can actually generate real images, I think it is not a problem for them to license Grok’s image generation pipeline

 

[19callum93]

Tim Cook/Apple has been far too interested in “Augmented Reality”, with AI now coming to bite them on the ass!