Apple Faces Criticism Over AI-Generated News Headline Summaries

ForkHandles said:
The BBC is the Premier news service in the UK. It’s the only news service I know of in the world that’s required, by law, to be impartial, balanced and factually correct.
The consequences of not being is being pilloried by government and by all of the billionaires that own other News outlets.
Click to expand...
I hope, sincerely, that you don't actually believe that and are just, in British parlance, taking a p!ss here
 
I don’t understand the point of summaries

Take something that takes 6 seconds to read, and make it take 4 seconds? But then, just in case it’s wrong or incomplete, read the full thing to be sure, so now your 6 second task took 10 seconds?


I’ve still yet to see a convincing demonstration on an actual, tangible, useful benefit to ANYTHING AI related.

Someone, please show me.
 
bluffalo said:
I don’t understand the point of summaries

Take something that takes 6 seconds to read, and make it take 4 seconds? But then, just in case it’s wrong or incomplete, read the full thing to be sure, so now your 6 second task took 10 seconds?


I’ve still yet to see a convincing demonstration on an actual, tangible, useful benefit to ANYTHING AI related.

Someone, please show me.
Click to expand...
I think that's accurate and agree that this phone stuff is just about the analog of random acts of mental violence on the user. Funny enough, I can actually give you an example of where it works but just not on a phone, and definitely NOT with any Apple Intelligence technology. I used Claude to help me rewrite an email to explain a complex ITIL topic, with the direction that it needed to be understood by someone that wasn't familiar with the subject matter. It worked surprisingly well and I edited here or there but left it largely intact. Something that would have taken me quite a while (I was capable but pressed for time as we swing the year closed at work) saved me a lot of typing and revising. What Apple offers is not what I did however. I also do not believe Apple Intelligence will ever be capable of this either without the escape hatch of something like GPT.
 
Have to say, so far AI has been a bust. Disabled it on my macs because there's no other way to stop it prompting useless suggestions in messages and mail. Just another example of how far behind they are.

To be fair though, I don't trust *any* LLM to not hallucinate.
 
IMG_9202.jpeg

At least Image Playground can create this image. Maybe Apple Intelligence is more self aware than I would have thought.

It’s not like bugs shouldn’t be expected, but the whole roll-out of their AI features sure seems less “Apple like” than I would have thought.
 
Last edited:
How long till it rewrites an article talking about Russia threatening to use nukes to Russia launching nukes and causing mass panic
 
Go to Settings > Apple Intelligence & Siri.

Near the top of the screen you'll see an "Apple Intelligence icon."

On the lower right-hand corner of the icon, you'll see a black "badge," similar to the red notification badges you see on apps with a number that indicates the number of notifications that haven't been "opened" in yet app.

And inside the black badge I mentioned in the beginning of the previous paragraph, you'll see the word "BETA."

This tells you that all features of Apple Intelligence are in beta.

Yes, even in the public release of iOS 18.2, Apple Intelligence itself is in beta.

So, Reporters Without Borders, or is it Reporters Without Brains? Leave Apple alone. Apple has made it abundantly clear that the feature is in beta, so mistakes can and will happen.
 
DrJR said:
The line for me was crossed when they stopped making printers. Sure I’ve bought every piece of hardware they’ve sold since then, but that was a line in the sand.
Click to expand...
There is no money to be made in low end printers. It only made sense when they had the best printers and no one had one that could print using Postscript in the small business market.
 
I don't see how this is any different that the false and baity headlines that go through media (mainstream or social) and readers don't bother to question. If you believe anything you read, that's your inability to think and verify that's the issue.
 
gsurf123 said:
There is no money to be made in low end printers. It only made sense when they had the best printers and no one had one that could print using Postscript in the small business market.
Click to expand...
Pretty sure there was a hint of sarcasm there friend
 
Mr_Ed said:
This crap’s just not ready for prime time. I still don’t understand the urge to push out these AI features few are actually asking for.
Click to expand...
I’m a lawyer and we have various AI tools and models available to us. All I can say that they are terrible for our use at the moment. When accuracy of words matter, these may as well produce random gibberish. That would actually be better as it would be easier to identify just how wrong the output is. I had high hopes for call transcripts and meeting summaries. But when nuance matters, these models can produce things that sound similar, but actually mean the exact opposite, or state the exact thing that shouldn’t be stated.

Probably the only good use I’ve found is a slightly better web search, where you then read the actual source material. But even then it can be completely and infuriatingly and stubbornly wrong at times.
 
podycust said:
Wow AI gets it wrong once and they want to burn the whole thing down?

If I recall correctly you have to select what notifications you would like to turn it on for and not on by default for everything??

Also BBC isn't the most stellar news source they try to make out they are!
Click to expand...
I mean, yes. If it’s stating completely incorrect things next to a masthead as if it’s news, it’s fairly easy to see why that’s a problem.
 
The rewrite feature is also not too accurate anymore, it makes up stories sometimes and I end up reverting the whole email or text to the original, it used to be better last month.
 
Apple is loosing its focus. IOS is extremely poor in features by now and the AI gimmick is not the savior.
Cant even correct a mistyped number in the dial pad!. Mail cant block or mark as spam senders, the way to edit and correct text is sooooo bad comapred to Android, etc.
 
Mr_Ed said:
This crap’s just not ready for prime time. I still don’t understand the urge to push out these AI features few are actually asking for.
Click to expand...
Because all of tech is hitting a plateau. There isn't enough new/novel experiences to give people with the existing technology we have. Then because companies are looking for the next big thing to try to drive continued gains (number must always go up!) they jump on the bandwagon of any little trend to try to make it the next big thing. The thing is, I don't think there is much left to do that's actually revolutionary within the bounds of the technology we have today. AI is just the current fad. I don't think most people care at all about AI features. I certainly don't.
 
alchemistmuffin said:
Apple should not comply. It’s under “Beta” itself and users have option to opt out of Apple Intelligence entirely to avoid this. Asking Apple to remove this feature is censorship in my opinion.
Click to expand...
A non-profit, non-governmental organization urging a corporation to remove a feature that attributes false information to a news organization is objectively not censorship.

The Apple Intelligence feature didn’t just give false information (which is already bad enough). The Apple Intelligence feature went further and specifically attributed that false information to BBC News (much, much worse).

This is not an issue to be brushed aside. For example, Romania had its recent presidential election result annulled due to Russian interference. Basically: fake news and false information has thrown a country of 19 million people into political turmoil.

To be clear: Apple Intelligence wasn’t the issue in Romania. Most of the problems in Romania happened on TikTok. Regardless, it’s not enough to say a feature is in “beta”. The spreading of false information (wherever it occurs) has serious, real-life consequences.
 
Chungry said:
I hope, sincerely, that you don't actually believe that and are just, in British parlance, taking a p!ss here
Click to expand...
Quite sincere.

If you can name a more reliable, non-partisan, source of news, I’m willing to hear it.

One of the great things about bbc news is how many people on the Left it annoys, exactly the same as the number of people on the Right. They’re usually the folk who live in the own echo chamber of news that chose to take issue with uncomfortable truths.

It’s it perfect? No!

Would I trust it over any of our newspapers or GBNews? In a nano second.
 
Last edited:
Apple doesn't have a clue of what to do with this new fancy "AI" technology.

Don't get me wrong, I love machine learning in many different forms and it is one of my hobbies, but this shoehorning LLM:s into where it's not at all needed is just ridiculous. I thought Apple was above that but apparently not.
 
I find Google's AI-generated search summaries very annoying too. Particularly as they take up so much space on the search results page, making you scroll down to skip past them. Needs to be an easy way to turn it off.
 
Clickbait headlines and partisan articles by human journalists and now AI summarising them with its own take. What a time to be alive.
 
Apple should have implemented true Artificial Intelligence. Apple Intelligence is just too bad since Apple can‘t jump.
 
I ended up turning it off because it would summarise messages to the exact opposite of what was actually said about 1/3 of the time.

Though sometimes it was so catastrophically wrong it made me laugh.

IMG_4819.jpeg
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Back
Top