Apple has said it will update, rather than pause, a new artificial intelligence (AI) feature that has generated inaccurate news alerts on its latest iPhones.
The company, in its first acknowledgement of the concerns, on Monday said it was working on a software change to "further clarify" when the notifications are summaries that have been generated by the Apple Intelligence system.
Beta software contains an implicit promise that the developer will actively work to squash bugs and make the product better before it goes final. Adding a warning label in the interim is an easy band-aid, but it doesn’t address the underlying problem. Apple needs to do much more work here, and if it can’t, it needs to turn this feature off until it can release a version it can stand behind.
I think it’s correct for Apple to leave this in control of users, not developers. But Apple really does need to make it more clear what is an AI-generated summary and what is a verbatim original notification. And, of course, it really needs to reduce the number of these errors.
In a statement to 9to5Mac today, Apple said it settled the case so it can “move forward from concerns about third-party grading that we already addressed in 2019.”
The company says that Siri data has “never been used to build marketing profiles, and it has never been sold to anyone for any purpose.”
The iPhone maker is considering adding new countries to the platform, which reaches about 125mn people every month in the US, Canada, UK and Australia, according to people familiar with the plans.
It was also considering building its locally focused news coverage in the UK, the people said, while adding its puzzles section in the country which is currently only available in the US and Canada.
Apple hopes that scaling up its News app, alongside work it is doing in improving advertising effectiveness, will help cement the tech group’s role as a growing source of revenues for many publishers.
According to Apple's release notes, iOS 18.2.1 addresses important bugs, and it is recommended for all users.
Apple today launched a new Shazam Fast Forward 2025 website that will spotlight 50 music artists who are poised to have a breakthrough year.
[...]
Artists are selected based on Shazam's "uniquely predictive data and algorithms," coupled with the "expertise of Apple Music's global editorial team."
Satechi's redesigned Stand & Hub brings enhanced NVMe SSD storage, faster SD card readers, and extra USB ports to Apple's redesigned Mac mini.
This means device makers won’t have to put their gadgets through a separate testing program for each platform to wear its “Works With” badge. If they get certified as a Matter Device by the CSA, they can show their results to the other ecosystems and get those badges, too, without doing any more testing.
Indonesia's has said Apple's plans to invest $1 billion in a manufacturing plant that produces components for smartphones and other products, details of which were still being ironed out.
As streaming cannibalizes traditional TV once and for all, it’s incorporating nostalgic formats and cobbling together all sorts of experiences to make sure you stay on one chosen app and that app alone—to demonstrate your brand loyalty in addition to your cultural enthusiasm. The future of streaming doesn’t just look like the past; it looks like a future in which you’re plugged in and engaged at all times via ever-increasing offers of exclusives that make one particular platform really stand out among all the others.
The way I see it, there are two related but differnet problems with the whole Apple Intelligence notification summarization things.
Firstly, obviously, the software is not working correctly all the time. We expected this from GenAI. I'm sure Apple is also expecting this, and in the Settings app, in small text, Apple said: "Summary accuracy may vary based on content."
I am skeptical that Apple can ever fix the summarization problems using GenAI. And I am not alone; a lot of people smarter than me also share this skeptism.
There is a different problem, though: Apple is putting misinformation into the mouths of other people. Okay, changing your mother's "That hike almost killed me" to "Attempted suicide but recovered" will cause panic, but you will probably laugh it off when you actually open the text messages and see what silly Apple has done.
But putting misinformation into the mouths of real journalistic organizations is a big no-no. Even though they can make real journalistic mistakes and mis-reports from time-to-time, organizations like BBC, I think, do take pride in getting things right and not misinform their audience. No matter how the software is labeled, "beta" or otherwise, any software that does this should not have been released to the general public.
(Yes, I am aware there is a little glyph next to the summarized notification. No one is going to know what that means, and no one should need to know what that means.)
So what Apple has promised to fix, so far, is to fix this second problem. Depending on how Apple updates the software, it may be enough for BBC to accept. But it will not solve the first problem, in that the software, the GenAI, is not good enough, and will never be good enough.
(Besides, what's the point to summarize a bunch of headlines of probably unrelated stories anyway?)
~
Severance, the Apple TV+ show, will be returning with the second season next week. I am now catching up, not by watching the episodes from season one, but by listening to the new podcast with a long title: The Severance Podcast with Ben Stiller & Adam Scott. The podcast is recapping the first season this week, episode-by-episode. I am still on the first episode, but it has been great so far. If you are a fan of the TV+ show, do take a listen.
~
Thanks for reading.