How to turn off Apple Intelligence on your iPhone
theverge.com
theverge.com
It's like humans calling the period they live in "contemporary". It make as little sense for stone age people to do it, as it did for medieval people and as it does for us. We need to start thinking more about the future and not the 10-years from now future but the 300 years from now future.
My prediction, based on no evidence or theory other than "the future will be like today, only in the future" is that AI will be called AI 300 years from now.
Do you feel better now?
That new AI tech might be branded as something else but I doubt it would be a problem.
ML is not a perfect term for marketing use, either. It is obscure, and to many people “learning” hints at some sort of human-like education process (though it is an industry term that does not imply anything of the sort).
I believe the most correct term to use in marketing and branding is… well, none. It is an implementation detail that presumably makes the product better at addressing certain problems or otherwise improve customer’s life. Talk about that instead.
really its easy to point a label at tangible visual stuff, but pointing a label at more ephemeral stuff creates all manners of issues.
First i’d be willing to say this following statement is true: ‘humans invented object identities and invented the labels, and also both the label machine and the labels pointing at identities they want to label’, well i guess darwin invented the machinery but humans willfully made the first use out of the machinery for that purpose. if humans choose to make labels and define where labels are applicable, theres never going to be an objective true valid standard sense of a word, except by convention but even that rests on ‘objective’ quicksand, unfortunately.
beyond an argument on linguistics and the philosophy of language, that still leaves the substances of intelligence itself. while llms have clear huge gaps in inadequacies, i suspect its not far off from human capabilities by algorithm adjustments, adding architecture modularly (like reasoning), and ultimately through distillation and transfer learning, plus new techniques. Because of that, even dumb haphazardly shoehorned llms might surprisingly be closer to intelligence (or a core component thereof) than what even the most ardent semantic nazis would consider ( or archnemesis to LLms, PhD, Gary Marcus, orrrr also famous naysayer and contrarian, Yann Lecunn).
300 years on the exponential singularity timescale might actually be 5 to 10 years.
but maybe it might be akin to TV standards of black before TVs became capable of true black. grey was black until oleds and now previous tv formats of black, aka really grey, can now retroactively be reclassified as grey. or how about usage of the words unlimited when it comes to cellular data. those naughty marketers and their definition bending usage of words.
But I'm curious, has anyone really made any effort to test this so-called AI first to see if it's at all useful, or lives up to any level of expectation?
Or is there some a priori evil element to justify this (Tim Cook slurping up all your data and using it for training and advertising without consent or opt out) that I don't know about?
All the promises of integrating deeply into the OS and have APIs directly into apps so you can use natural language to get any app to do stuff for you is vaporware
So much potential but none of it delivered yet, I hope it will change soon
Here is the best commentary that I've read on the subject from someone who works regularly in generative AI: https://xeiaso.net/blog/2025/squandered-holy-grail/
In iOS, if you search in settings for something in the realm of Apple Intelligence. The results will be dominated by Apple Intelligence results, which is a good chance that that's not what you want. On Amazon.com, they threw their entire review history into Rufus, and the results are often are just plain outdated and wrong. It's still garbage in, garbage out with generative AI, and these companies have completely ignored that.
Hold on, because there's a ton of money riding on massive adoption, whether you want to adopt it or not.
Early adopters should recognize first versions of forward-looking tech are going to be less-than-optimal, and that the warts will eventually fall off.
Y’all around here advocate for shipping whatever you have, but when a larger org applies your principles, everyone is in a tizzy.
I would be surprised if most iPhone users couldn't figure this out themselves so my personal hunch is this article is mostly clickbait for anti-AI folks who already have these features disabled. In particular, because the only way to disable it is to have opted in to it in the first place and that flow is very similar to the disable flow.
Let alone reading what’s on their screen when they enable features, promptly forgetting they enabled it, thus asking how to disable them…
Ok enough. Have a good day :)
They dont need to group the titles together like that, it feels deceptive. I feel like the average user will be put off disabling AI if they still want to use Siri like they always have.
You and I might be knowledgeable enough to realize that we can turn such options off, but the vast, vast, vast majority of people are not.
Sure. Lots of people are too tech-unsavvy to figure that out. But those same people won't be googling how to do it. They'll just give up and ask their kid to do it for them or something.