...show you AI topless nudes of Taylor Swift nobody* asked for.
*other than probably Musk himself, on his usual 3am bender
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
iOS 26 to Bring ChatGPT-5 Integration to Apple Intelligence
- Thread starter MacRumors
- Start date
- Sort by reaction score
...show you AI topless nudes of Taylor Swift nobody* asked for.
*other than probably Musk himself, on his usual 3am bender
Yes this is an advantage of Grok, it has realtime access to the X API.
Although there are a ton of issues with model alignment I think it has a strong future.
Particularly since xAI is targeting “adult” use and given Elon’s funding and the API / scale advantage I think they will be a strong competitor in the years to come. Adult content always pushes technology forward in adoption, from vhs to blu ray to internet service, for better or worse.
Except X just lost its CEO, has not become the super app for crypto/payments/banking/ecommerce that Musk promised, gets advertising contracts through coercion, and got rid of moderation and became "anti-woke" (read "far right"), losing a lot of users. It does not look good for X as a platform. If xAI has good search results for current news, how long will that last if X continues the way it has been going? How long before those news results start to sound just as biased as Fox News or just become "alternative facts"?
Status of scraping copyrighted material?
Status of all Lawsuits about scraped material?
When do copyrighted material world views change?
LLMs and scraping copyrighted Material.
Liabilities of LLMs scraped materials.
So many unknowns.
Status of all Lawsuits about scraped material?
When do copyrighted material world views change?
LLMs and scraping copyrighted Material.
Liabilities of LLMs scraped materials.
So many unknowns.
You will have to wait for the industry to standardize APIs and feature sets before you get to set your default LLM. You might also have to wait for competition to eliminate enough companies so that governments have a pretext to step in and demand Apple to allow changing the default. Honestly, it is too early because these integrations are going to be lame for a year or two. Plus, you cannot trust Apple to do anything right with Siri given its history. For now, you can get apps or use the web to get any chatbot integration on your iPhone.
From a scale standpoint, almost none of that matters. Enough people still use X / Twitter that there is a relatively valuable stream of data especially compared to what others have to pay for or don't have such realtime integration with. I think Meta is the only other one who can compete in that specific area.
They already sound biased, Grok especially with personas is a disaster but a portion of the public will still use it. If it gets too extreme they'll dial it back. Grok was #1 on the App store a couple weeks ago for a few days, and some gimmick hook is all it will take to get there again.
Elon worked with Nvidia to scale out an entire datacenter build in 19 days from first arrival to model training this summer. No other company is operating at even 20% that speed. Eventually it will pay some dividends.
I'm not saying it's the best model, or even a necessarily a good one given all of the concerns, but people who write it off entirely because of "mechahitler" kind of nonsense (which, yes that was obviously terrible) are missing the point.
Way, way too many people focus on the drama and clickbait outrage that we're all supposed to eat up every day and obsess over. There are real issues, I"m not discounting that, but they don't really matter in the long-term. Elon is still going to forge ahead, so will Mark Zuckerberg, so will Google, so will Sam Altman, etc.
The Eleanor Roosevelt quote:
"Great minds discuss ideas, average minds discuss events, small minds discuss people" is extremely applicable to so many facets of modern life. Virtually all of the discourse and social media is flooded with the small minded and average minded points of view, few really see the long-term.
I say none of this defending these people, I have issues with them too, but I don't write off their products or the impacts they'll have because I disagree with the people running them. Facebook has been transformative, mostly in a bad way, and I don't use that platform at all. It's an objective fact, no matter how much I post about how terrible I think Mark is or genocides in Myanmar or whatever. None of it gets better telling other people at our level how terrible things are all it does is distract us from the paradigm shift of society happening whether we like it or not. It's best to be informed, from my point of view, not focused on clickbait nonsense that is ultimately trivial.
That said, people who really go out in public and actually protest, start competing organizations, etc. and operate in the real world not online have my utmost respect. The people that only do it on the internet are having zero impact. Society is a mess, these tools will probably mess it up more, and you telling me that or me telling other people that won't change a thing.
edit: you can see the above play out this very thread, especially if you look at the post history of the people who frame their entire online engagements in the way described. It’s bizarre, I don’t get it, but a lot of people love to engage that way… it’s a shame.
Last edited:
ChatGPT (and some other LLMs?) sometimes has a problem with political questions, even just questions about voting laws. Sometimes it'll answer them with no argument, and other times it'll refuse, saying something to the effect that "I can't get into politics". When that happens, usually all I've needed to do is tell it "Yes you can", though sometimes I have to get a little more insistent.
Last edited:
Usually you can prompt ChatGPT to look online for current info by asking it something like "What's the current situation on ...?" So far, for me, using the word "current" has been all it takes, and usually you'll then see it display "Searching the web..."
Last edited:
Where does it say ChatGPT 5 will only be in iOS 26? I'm using it under iOS 18 right now. OpenAI is no longer offering versions 4 and 4.5 to anyone.Except it doesn't make sense.
"With the rollout of GPT-5, OpenAI is getting rid of the model picker and retiring many of its older models, including GPT-4o, GPT-4.1, and GPT-4.5."
There is no reason it shouldn't be 5 on everything since the old models are supposedly gone now.
Attachments
what free AI app we using on our phones? i really dont want another subscription but ChatGPT locks me out after just a few queries now. i have CoPilot and gemini and claude installed too.
As far as I can tell CoPilot doesn't lock you out and seems to be good at most of the tasks I throw at it.
As far as I can tell CoPilot doesn't lock you out and seems to be good at most of the tasks I throw at it.
Grok has gotten rave reviews from Hitler, Goebbels, Himmler, Mengele, Braun and the entire RepubliKKKan party.
… and i predict soon after iPh 17’s roll out, there will be glass action lawsuits …
Any from of enchantment is great since it mostly a chat bot at the moment but would really be great if Apple can up their game on their intelligent side so that thing like image playground or Siri can be better or at least up to expectation. They must for go the pass way of updating Siri yearly or only by iOS update they should do it in the background just like how Google or even Grok is doing it. Hopefully they can do it!
I'm being really cautious on this stuff. I don't trust these AI companies at all. Anything you say will be used against you at a later date.
Oh hell, they're gonna know I'm interested in Carnitas recipes and high performance computing!
There are tradeoffs to anything, I agree regarding things like using it for therapy etc. which is a horrible use case in general IMO, but there is a whole lot of utility that really doesn't get into anything I'd be worried about someone knowing.
I have an entire apparatus to stop tracking me around because I hate targeted ads and profiling, but I still feel this way about GenAI in general – there's utility there and right now they aren't selling my data and there's nothing in there that a subpoena could reveal that would be damning to me.
To each their own though, everyone has their own risk profile. Unless you take enormous steps (I do) everything you do online is deeply tracked. It's like a part time job to stop that from happening, truly. I do it, but the 'normal' people in my life absolutely hate it because it breaks a lot of functionality for junk websites they're used to. I just ask them to use cellular when they need to access those things and it works out well enough.
The realID stuff that is happening with the general internet and laws against encryption is where we really need to focus efforts on fighting back at the moment, but there's a sea change going on with that I'm not sure we'll be able to escape unfortunately.
