Apple and OpenAI allegedly reach deal to bring ChatGPT functionality to iOS 18

Posted:
in iOS edited May 29

Perhaps against the wishes of Apple's AI chief John Giannandrea, Apple and OpenAI have reportedly reached a deal that will bring ChatGPT to iOS 18 and other operating systems in 2024.

An image combining Siri with the ChatGPT logo
Apple and OpenAI could bring ChatGPT functionality to Siri



Apple wants to bring AI to its platforms, but it has focused on privacy-preserving on-device models and not the Large Language Models (LLMs) used by its competitors. So, the company has been seeking a partnership with Google or Microsoft to bring their LLMs to iPhone.

According to a report, a deal has finally been struck between OpenAI and Apple. The details of the deal aren't known, but it likely means technologies used for ChatGPT could be accessible by Apple users, namely through Siri.

A deal with OpenAI doesn't close out options with other AI industry players. One rumor suggested Apple was looking to create an AI App Store, so OpenAI could be one of several companies populating such a store.

Internal politics are the likely driver of this compromise, as the report shared that Apple's Senior Vice President Machine Learning and AI Strategy, John Giannandrea, said in an email in 2023 that "the last thing people needed was another chatbot." So, either Apple didn't bother with a server-side LLM or they started too late and it isn't ready.

Whatever the case, if this report is true, Apple likely won't be announcing an Apple-made LLM to replace Siri. Instead, it will use server-side products like ChatGPT to answer user questions.

For example, today if you ask Siri something that would require a search, Siri sends a search query to Google using an anonymous identifier. OpenAI could have a similar option for users where queries requiring AI would pass information to ChatGPT in a privacy-preserving way and return a useful answer read by Siri.

The WWDC keynote where everything will be revealed is on June 10. Stay tuned to AppleInsider as we continue to break exclusive leaks surrounding iOS 18 and AI.

Rumor Score: Possible

Read on AppleInsider

«1

Comments

  • Reply 1 of 37
    tmaytmay Posts: 6,428member
    Pearl clutching, and handwringing, ensue...
    40domiCrossPlatformFroggerwatto_cobra
  • Reply 3 of 37
    danoxdanox Posts: 3,076member
    We shall see I anticipate Apple will temporarily use OpenAI like they used Google for Maps until they get their version online.......Note the Google geeks are still up in arms about it like many in tech are still mad about Apple dumping Intel keeping AMD and Nvidia at arms length with their SOC designs.

    https://www.reddit.com/r/hardware/comments/1cwdlak/notebookcheck_apple_m4_soc_analysis_amd_intel_and/ Geek Meltdown :smile: 
    edited May 29 Alex1Nwatto_cobra
  • Reply 4 of 37
    slow n easyslow n easy Posts: 361member
    This is very welcome news for me. I don't even try to ask Siri anything even remotely complicated anymore. All I use Siri for is to turn on or off my lights and other Apple related things and simple stuff like weather and sports scores. Like for example today I wanted to know "from which country is Ulanzi based in" and I already knew that there would be zero chance of Siri having any kind of answer. I asked ChatGPT and got a definite answer that I feel very confident is accurate.
    ssfe11williamlondonAlex1N
  • Reply 5 of 37
    princeprince Posts: 90member
    but but but DED was frothing at the mouth that nothing can compete with Apple AI, especially not CoPilot.
    And now Apple is putting OpenAI, that is part of the power behind CoPilot, in their phones.

    DED swings, and misses as usual.
    Bing Chat isn't OpenAI. Also, Microsoft's Copilot PC isn't really about AI or "Apple's AI," but about an effort by Microsoft to raise the price of the PC to where Apple is selling its Macs, rather than in the bargain basement of generic Android/PCs. That's something Microsoft has been really bad at doing, as the article you referenced made clear--ever since it started trying to do this in the iPod days. 

    If you don't understand an article you just read, maybe you should hesitate to put your fake chat name on your arrogant opinions about it, because like the thousands of hecklers over the last 15 years, you'll end up looking like another bitter crank who ended up with egg on his face trying to scramble up a hater omelette.
    40domiradarthekatbadmonkssfe11tmayAlex1Njas99watto_cobraroundaboutnowBart Y
  • Reply 6 of 37
    gatorguygatorguy Posts: 24,382member
    danox said:
    We shall see I anticipate Apple will temporarily use OpenAI like they used Google for Maps until they get their version online.......Note the Google geeks are still up in arms about it like many in tech are still mad about Apple dumping Intel keeping AMD and Nvidia at arms length with their SOC designs.

    https://www.reddit.com/r/hardware/comments/1cwdlak/notebookcheck_apple_m4_soc_analysis_amd_intel_and/ Geek Meltdown :smile: 
    Ahh, the Qualcomm is doomed, Intel is doomed discussion. :)  
    ctt_zh
  • Reply 7 of 37
    40domi40domi Posts: 138member
    Not a good move by Apple, it clearly sends a signal they're behind 😏
    williamlondon
  • Reply 8 of 37
    ctt_zhctt_zh Posts: 83member
    prince said:
    but but but DED was frothing at the mouth that nothing can compete with Apple AI, especially not CoPilot.
    And now Apple is putting OpenAI, that is part of the power behind CoPilot, in their phones.

    DED swings, and misses as usual.
    Bing Chat isn't OpenAI. Also, Microsoft's Copilot PC isn't really about AI or "Apple's AI," but about an effort by Microsoft to raise the price of the PC to where Apple is selling its Macs, rather than in the bargain basement of generic Android/PCs. That's something Microsoft has been really bad at doing, as the article you referenced made clear--ever since it started trying to do this in the iPod days. 

    If you don't understand an article you just read, maybe you should hesitate to put your fake chat name on your arrogant opinions about it, because like the thousands of hecklers over the last 15 years, you'll end up looking like another bitter crank who ended up with egg on his face trying to scramble up a hater omelette.
    Bing Chat (now called Copilot) is based on OpenAI. Below is part of a Microsoft post from well over a year ago. The following link has some further notes on Copilot / OpenAI's ChatGPT....

    Copilot (formerly Bing Chat) vs. ChatGPT: Which AI chatbot should you use? (zapier.com) 

    Do you have proof that Microsoft's Copilot PC isn't about AI but is an effort to raise the price of the PC to where Apple is selling its Macs? The Microsoft Store price for a 15" 16GB/512GB 12th Gen Intel Core i7 machine was £1799 in the UK, whereas the Microsoft Store price for a Snapdragon X Elite 12-core 15" machine is £1549 (again for the 16GB/512GB variant).. a cool saving of £250. Good of Microsoft to pass on the savings, perhaps not every manufacturer has / will....


    edited May 30 CrossPlatformFrogger
    bing.jpg 183.2K
  • Reply 9 of 37
    avon b7avon b7 Posts: 7,854member
    danox said:
    We shall see I anticipate Apple will temporarily use OpenAI like they used Google for Maps until they get their version online.......Note the Google geeks are still up in arms about it like many in tech are still mad about Apple dumping Intel keeping AMD and Nvidia at arms length with their SOC designs.

    https://www.reddit.com/r/hardware/comments/1cwdlak/notebookcheck_apple_m4_soc_analysis_amd_intel_and/ Geek Meltdown :smile: 
    I suppose that is a tacit admission that you accept Apple is behind here. 

    If you think Apple, two years after the buzz around this subject hit the mainstream, will still have to 'temporarily' use a competitor's solution, then it doesn't look great. 

    Most people up to now have simply pointed to the painfully obvious fact that Apple has been behind because it had no competing solution. 

    It was always unwise to counterargue that it wasn't behind and pointing to the future. 

    Those competitors also have a future and are moving very, very fast. 

    We just don't know what will be announced next month and also have no idea when it will reach Apple users. I suppose we are realistically looking at year end.

    The question then will be 'in what state?' Beta? Fully baked? Fully baked but limited? Will they try to monetise it? If so, to what degree.

    If you have the possibility of integrating someone else's solution, it's nice to have those options while you work on your own. Some people will have problem with that. 

    I very much doubt Apple even considered going another year without these options, so at some point over these last two years someone probably said it's time to pivot the ship to AI. Late but better late than never. 

    At this point, it's clear to me that Apple got caught off guard with this. 

    Better to accept that, move on and look forward to having those options available at system level. 


    ctt_zhmuthuk_vanalingamAlex1N
  • Reply 10 of 37
    MacProMacPro Posts: 19,778member
    If Apple announces at WWDC, as expected, they can hardly be 'just coming to an arrangement' with anyone.  This has to have been baked in for a while.  

    Apple's advantage is getting AI into end users' hands and increasing the sales potential for device upgrades.  Most other companies don't have the same level of market penetration of end-user products with the potential to run AI locally. This is going to be huge over the next few years, and I can't wait to see where AAPL goes.

    I just read that Apple News+ has been astronomically successful, yet another area in which Apple is 'new' and 'late to the game'  then takes an industry lead. Do I really need to list all the other examples?  When will the nay-sayers get it?  Apple tends to succeed (to put it mildly), even if not immediately.
    danoxAlex1Njas99watto_cobraBart Y
  • Reply 11 of 37
    MacProMacPro Posts: 19,778member

    avon b7 said:
    danox said:
    We shall see I anticipate Apple will temporarily use OpenAI like they used Google for Maps until they get their version online.......Note the Google geeks are still up in arms about it like many in tech are still mad about Apple dumping Intel keeping AMD and Nvidia at arms length with their SOC designs.

    https://www.reddit.com/r/hardware/comments/1cwdlak/notebookcheck_apple_m4_soc_analysis_amd_intel_and/ Geek Meltdown :smile: 
    I suppose that is a tacit admission that you accept Apple is behind here. 

    If you think Apple, two years after the buzz around this subject hit the mainstream, will still have to 'temporarily' use a competitor's solution, then it doesn't look great. 

    Most people up to now have simply pointed to the painfully obvious fact that Apple has been behind because it had no competing solution. 

    It was always unwise to counterargue that it wasn't behind and pointing to the future. 

    Those competitors also have a future and are moving very, very fast. 

    We just don't know what will be announced next month and also have no idea when it will reach Apple users. I suppose we are realistically looking at year end.

    The question then will be 'in what state?' Beta? Fully baked? Fully baked but limited? Will they try to monetise it? If so, to what degree.

    If you have the possibility of integrating someone else's solution, it's nice to have those options while you work on your own. Some people will have problem with that. 

    I very much doubt Apple even considered going another year without these options, so at some point over these last two years someone probably said it's time to pivot the ship to AI. Late but better late than never. 

    At this point, it's clear to me that Apple got caught off guard with this. 

    Better to accept that, move on and look forward to having those options available at system level. 


    This isn't about being 'behind' it's about 'how' and 'when' to have the best results and long-term benefits to Apple and the billions of Apple device users.  Not about 'I got there. First, I have a website where people using computer products can play.'
    edited May 30 tmayAlex1Njas99watto_cobraBart Y
  • Reply 12 of 37
    MacProMacPro Posts: 19,778member

    gatorguy said:
    danox said:
    We shall see I anticipate Apple will temporarily use OpenAI like they used Google for Maps until they get their version online.......Note the Google geeks are still up in arms about it like many in tech are still mad about Apple dumping Intel keeping AMD and Nvidia at arms length with their SOC designs.

    https://www.reddit.com/r/hardware/comments/1cwdlak/notebookcheck_apple_m4_soc_analysis_amd_intel_and/ Geek Meltdown :smile: 
    Ahh, the Qualcomm is doomed, Intel is doomed discussion. :)  
    I think they will all do very well. Apple certainly seems well-positioned to benefit from a mature, well-thought-out strategy.
    tmayAlex1Nwatto_cobraroundaboutnowBart Y
  • Reply 13 of 37
    So Siri is so stupid to be fixed internally by Apple or the reverse. 

    What does this cooperation mean? Is it cloud-based AI or on device AI? 

    Will OpenAI get the access to user data?
    nubus
  • Reply 14 of 37
    MacProMacPro Posts: 19,778member
    So Siri is so stupid to be fixed internally by Apple or the reverse. 

    What does this cooperation mean? Is it cloud-based AI or on device AI? 

    Will OpenAI get the access to user data?
    Regarding the last question, do you have to ask?  Come on.
    Alex1Nwatto_cobraroundaboutnowBart Y
  • Reply 15 of 37
    avon b7avon b7 Posts: 7,854member
    MacPro said:

    avon b7 said:
    danox said:
    We shall see I anticipate Apple will temporarily use OpenAI like they used Google for Maps until they get their version online.......Note the Google geeks are still up in arms about it like many in tech are still mad about Apple dumping Intel keeping AMD and Nvidia at arms length with their SOC designs.

    https://www.reddit.com/r/hardware/comments/1cwdlak/notebookcheck_apple_m4_soc_analysis_amd_intel_and/ Geek Meltdown :smile: 
    I suppose that is a tacit admission that you accept Apple is behind here. 

    If you think Apple, two years after the buzz around this subject hit the mainstream, will still have to 'temporarily' use a competitor's solution, then it doesn't look great. 

    Most people up to now have simply pointed to the painfully obvious fact that Apple has been behind because it had no competing solution. 

    It was always unwise to counterargue that it wasn't behind and pointing to the future. 

    Those competitors also have a future and are moving very, very fast. 

    We just don't know what will be announced next month and also have no idea when it will reach Apple users. I suppose we are realistically looking at year end.

    The question then will be 'in what state?' Beta? Fully baked? Fully baked but limited? Will they try to monetise it? If so, to what degree.

    If you have the possibility of integrating someone else's solution, it's nice to have those options while you work on your own. Some people will have problem with that. 

    I very much doubt Apple even considered going another year without these options, so at some point over these last two years someone probably said it's time to pivot the ship to AI. Late but better late than never. 

    At this point, it's clear to me that Apple got caught off guard with this. 

    Better to accept that, move on and look forward to having those options available at system level. 


    This isn't about being 'behind' it's about 'how' and 'when' to have the best results and long-term benefits to Apple and the billions of Apple device users.  Not about 'I got there. First, I have a website where people using computer products can play.'
    Oh no. I am fully aware of what's out there and have commented on it many, many times here. Just do a search here on my posts with Ascend, Mindspore, Cann or Pangu etc.

    The difference over the last two years has been how quickly some solutions have reached 'userland' (directly or indirectly) with little to no effort on the user's part and how blazingly fast those solutions have evolved.

    When things like Sora get a universal release it will have to come with some very tight reigns. 

    End users won't be interested in running weather prediction models or models for pharmaceutical research. They don't need to know about NLP, NLU, NLG even though they are touched by it every day. 

    But when AI solutions using these models are used daily from within universal apps like WhatsApp and use natural language etc to interact, that is where a lot of the buzz has been over the last two years. 

    A buzz that, with hindsight, Apple probably amplified by refusing to even utter the words when the spotlight fell on it. 

    Being behind is a statement of fact. No more. No less. 

    Strategically, Apple got caught short. What comes next is anybody's guess but this article points to 'outside' solutions. I'm saying that shouldn't be a issue for anyone but obviously some of those who have claimed Apple isn't behind will have issues with it.

    If the Flintstones had access to these options, no doubt Siri would be a good fit for the times and it doesn't surprise me that its due for a huge AI overhaul (according to rumours). However, one of Apple's main Siri issues seems to be that there isn't one Siri for all devices but multiple Siris, all performing differently. 
    edited May 30 ctt_zh
  • Reply 16 of 37
    MacProMacPro Posts: 19,778member
    avon b7 said:
    MacPro said:

    avon b7 said:
    danox said:
    We shall see I anticipate Apple will temporarily use OpenAI like they used Google for Maps until they get their version online.......Note the Google geeks are still up in arms about it like many in tech are still mad about Apple dumping Intel keeping AMD and Nvidia at arms length with their SOC designs.

    https://www.reddit.com/r/hardware/comments/1cwdlak/notebookcheck_apple_m4_soc_analysis_amd_intel_and/ Geek Meltdown :smile: 
    I suppose that is a tacit admission that you accept Apple is behind here. 

    If you think Apple, two years after the buzz around this subject hit the mainstream, will still have to 'temporarily' use a competitor's solution, then it doesn't look great. 

    Most people up to now have simply pointed to the painfully obvious fact that Apple has been behind because it had no competing solution. 

    It was always unwise to counterargue that it wasn't behind and pointing to the future. 

    Those competitors also have a future and are moving very, very fast. 

    We just don't know what will be announced next month and also have no idea when it will reach Apple users. I suppose we are realistically looking at year end.

    The question then will be 'in what state?' Beta? Fully baked? Fully baked but limited? Will they try to monetise it? If so, to what degree.

    If you have the possibility of integrating someone else's solution, it's nice to have those options while you work on your own. Some people will have problem with that. 

    I very much doubt Apple even considered going another year without these options, so at some point over these last two years someone probably said it's time to pivot the ship to AI. Late but better late than never. 

    At this point, it's clear to me that Apple got caught off guard with this. 

    Better to accept that, move on and look forward to having those options available at system level. 


    This isn't about being 'behind' it's about 'how' and 'when' to have the best results and long-term benefits to Apple and the billions of Apple device users.  Not about 'I got there. First, I have a website where people using computer products can play.'
    Oh no. I am fully aware of what's out there and have commented on it many, many times here. Just do a search here on my posts with Ascend, Mindspore, Cann or Pangu etc.

    The difference over the last two years has been how quickly some solutions have reached 'userland' (directly or indirectly) with little to no effort on the user's part and how blazingly fast those solutions have evolved.

    When things like Sora get a universal release it will have to come with some very tight reigns. 

    End users won't be interested in running weather prediction models or models for pharmaceutical research. They don't need to know about NLP, NLU, NLG even though they are touched by it every day. 

    But when AI solutions using these models are used daily from within universal apps like WhatsApp and use natural language etc to interact, that is where a lot of the buzz has been over the last two years. 

    A buzz that, with hindsight, Apple probably amplified by refusing to even utter the words when the spotlight fell on it. 

    Being behind is a statement of fact. No more. No less. 

    Strategically, Apple got caught short. What comes next is anybody's guess but this article points to 'outside' solutions. I'm saying that shouldn't be a issue for anyone but obviously some of those who have claimed Apple isn't behind will have issues with it.

    If the Flintstones had access to these options, no doubt Siri would be a good fit for the times and it doesn't surprise me that its due for a huge AI overhaul (according to rumours). However, one of Apple's main Siri issues seems to be that there isn't one Siri for all devices but multiple Siris, all performing differently. 
    We'll see how it pans out in a few years. Similar sentiments were prevalent at the launch of many things Apple got into 'late' or 'not knowledgeable enough about.'  I have made millions not agreeing with those sentiments so far. I do not disagree that Siri is neither good enough nor cross-device consistent enough at present.
    edited May 30 tmaydanoxAlex1Nwatto_cobraroundaboutnow
  • Reply 17 of 37
    tmaytmay Posts: 6,428member
    avon b7 said:
    MacPro said:

    avon b7 said:
    danox said:
    We shall see I anticipate Apple will temporarily use OpenAI like they used Google for Maps until they get their version online.......Note the Google geeks are still up in arms about it like many in tech are still mad about Apple dumping Intel keeping AMD and Nvidia at arms length with their SOC designs.

    https://www.reddit.com/r/hardware/comments/1cwdlak/notebookcheck_apple_m4_soc_analysis_amd_intel_and/ Geek Meltdown :smile: 
    I suppose that is a tacit admission that you accept Apple is behind here. 

    If you think Apple, two years after the buzz around this subject hit the mainstream, will still have to 'temporarily' use a competitor's solution, then it doesn't look great. 

    Most people up to now have simply pointed to the painfully obvious fact that Apple has been behind because it had no competing solution. 

    It was always unwise to counterargue that it wasn't behind and pointing to the future. 

    Those competitors also have a future and are moving very, very fast. 

    We just don't know what will be announced next month and also have no idea when it will reach Apple users. I suppose we are realistically looking at year end.

    The question then will be 'in what state?' Beta? Fully baked? Fully baked but limited? Will they try to monetise it? If so, to what degree.

    If you have the possibility of integrating someone else's solution, it's nice to have those options while you work on your own. Some people will have problem with that. 

    I very much doubt Apple even considered going another year without these options, so at some point over these last two years someone probably said it's time to pivot the ship to AI. Late but better late than never. 

    At this point, it's clear to me that Apple got caught off guard with this. 

    Better to accept that, move on and look forward to having those options available at system level. 


    This isn't about being 'behind' it's about 'how' and 'when' to have the best results and long-term benefits to Apple and the billions of Apple device users.  Not about 'I got there. First, I have a website where people using computer products can play.'
    Oh no. I am fully aware of what's out there and have commented on it many, many times here. Just do a search here on my posts with Ascend, Mindspore, Cann or Pangu etc.

    The difference over the last two years has been how quickly some solutions have reached 'userland' (directly or indirectly) with little to no effort on the user's part and how blazingly fast those solutions have evolved.

    When things like Sora get a universal release it will have to come with some very tight reigns. 

    End users won't be interested in running weather prediction models or models for pharmaceutical research. They don't need to know about NLP, NLU, NLG even though they are touched by it every day. 

    But when AI solutions using these models are used daily from within universal apps like WhatsApp and use natural language etc to interact, that is where a lot of the buzz has been over the last two years. 

    A buzz that, with hindsight, Apple probably amplified by refusing to even utter the words when the spotlight fell on it. 

    Being behind is a statement of fact. No more. No less. 

    Strategically, Apple got caught short. What comes next is anybody's guess but this article points to 'outside' solutions. I'm saying that shouldn't be a issue for anyone but obviously some of those who have claimed Apple isn't behind will have issues with it.

    If the Flintstones had access to these options, no doubt Siri would be a good fit for the times and it doesn't surprise me that its due for a huge AI overhaul (according to rumours). However, one of Apple's main Siri issues seems to be that there isn't one Siri for all devices but multiple Siris, all performing differently. 
    Unless it actually effects Apple sales, then being "late" or "behind", for integrating some 3rd party solution, is immaterial. You have made many similar predictions in the past, almost all of which ultimately have had zero impact on Apple sales. 

    I for one welcome all of the semiconductor and hardware companies in the computer business today, that are now roughly technology equals, something that hasn't been the case for many years. This ultimately leads to more competition and more choice. 

    If AI is really a differentiator in products, then that should become apparent pretty early. My own opinion is that there will be a strong increase in hardware sales, and likely little shift in platform share. The big question in my mind is whether x86 can survive the onslaught of Arm.
    Alex1Nwatto_cobraroundaboutnowBart Y
  • Reply 18 of 37
    Even IF this is true, so what?  Siri doesn’t generate revenue, and Apple will not compromise security or privacy with any partnerships, so what would be the harm?  I really don’t care either way.  Offering different LLMs in the App Store might be interesting though.
    Alex1Njas99watto_cobra
  • Reply 19 of 37
    danoxdanox Posts: 3,076member
    gatorguy said:
    danox said:
    We shall see I anticipate Apple will temporarily use OpenAI like they used Google for Maps until they get their version online.......Note the Google geeks are still up in arms about it like many in tech are still mad about Apple dumping Intel keeping AMD and Nvidia at arms length with their SOC designs.

    https://www.reddit.com/r/hardware/comments/1cwdlak/notebookcheck_apple_m4_soc_analysis_amd_intel_and/ Geek Meltdown :smile: 
    Ahh, the Qualcomm is doomed, Intel is doomed discussion. :)  
    Intel and AMD are doomed if they don't get the wattage down and the performance up when unplugged from a wall socket, Qualcomm is making small steps in the right direction unfortunately they are dependent on Microsoft for a OS and they are a bit of a patent troll which means there will be a Recall on the marriage.
    Alex1Njas99watto_cobra
  • Reply 20 of 37
    CheeseFreezeCheeseFreeze Posts: 1,291member
    Highly confusing messaging around AI as of late. It looks like Apple would get a smorgasbord of LLM models, which doesn’t “feel” very Apple.

    We’ll see what it all means soon! 
    williamlondonAlex1N9secondkox2
Sign In or Register to comment.