Underwhelming performance of Apple Intelligence will hit iPhone sales, Kuo claims

I doubt it. Apple Intelligence is a feature in development, not a selling point. People will continue to buy iPhones for the reasons they bought the last one they bought.

That said, regardless of your “feelings” about Apple Intelligence one way or another, I can assure that nearly EVERY person with access to the writing tools should make more use of at least that aspect of AI.

For me, the Genmoji/Image Generator will be the part I use the least, but see that’s the great thing about Apple Intelligence — you can use what proves useful to you, and ignore what isn’t useful to you. Gosh, it’s almost like they designed it to work like that.

By now it's clear that Apple Intelligence (and other AI) isn't really driving smartphone hardware sales. There was no big pop in sales when the Apple Intelligence capable iPhone 16 family debuted. So a slow rollout of Apple Intelligence features likely won't impact iPhone sales greatly.

At least in early 2025 consumer-facing artificial intelligence (in general, not just Apple's offering) is still essentially alpha software. Even Apple labels Apple Intelligence with a "Beta" badge but it's barely that (look at the news headlines summary debacle).

Clearly you don't need Apple Intelligence for AI assisted writing tools, the third party AI offerings are perfectly capable of that.

It's worth noting that Mac sales aren't really affected by Apple Intelligence either; the macOS implementation trails the iOS implementation.

Here are the things that consumers prioritize in smartphones: cameras, display, battery life, and price. It's not rocket science. Not peak download speeds on mmWave 5G networks. Not how many polygons it can paint on the screen, not how many TFlops of calculations, not Geekbench scores.

I still believe that AI has great potential in the enterprise market and I know it has been used by some very large corporations over several years now. But a context-aware AI chatbot assistant is still years away based on my experiences with 7-8 chatbots. Those tools have way too many mistakes to make them trustworthy so the user ends up spending a lot more time double or triple checking their output. These chatbots aren't trustworthy. They will make up falsehoods and pass them off as fact without any sense of remorse or hesitation.

One piss poor tendency that AI designers have programmed into models is the inability for an AI assistant to understand when a question is beyond its capabilities. Trust is earned and a lot of LLM-powered AI chatbot assistants squander some of that trust every time they hallucinate or shovel out some B.S. as an answer.

Sure, the engineers say "we can fix that" but what they can't fix is the loss of trust. Some of them are belatedly understanding this and now adding more and more boilerplate to answers saying "okay, we might be wrong so please double check the answer." That's right, consumers are being asked to do debugging for piss poor LLMs. But for some of us, it's already too late. That trust ship sailed months ago.

I think a handful of Apple executives are realizing this. Half-assed AI will ultimately be more damaging than a delayed version of something actually useful and trustworthy.

My guess is within five years, 75% of these AI companies will have disappeared from the landscape. Some of the technology will likely be absorbed by some of the giants for internal use on proprietary tasks. But a large number of these consumer-facing AI services will be laughable memories.

And in five years a lot of AI engineers will be changing careers because they won't get hired by another company. AI will be writing most of the code. The people who get AI right won't want to hire the people who got it wrong. This is a critical time for Apple. They have to figure out if the right person is driving the stagecoach right now. They can't afford to keep passing the reins over to someone else or it'll end up like Project Titan (Apple Car).

chasm said:

I doubt it. Apple Intelligence is a feature in development, not a selling point. People will continue to buy iPhones for the reasons they bought the last one they bought.

That said, regardless of your “feelings” about Apple Intelligence one way or another, I can assure that nearly EVERY person with access to the writing tools should make more use of at least that aspect of AI.

For me, the Genmoji/Image Generator will be the part I use the least, but see that’s the great thing about Apple Intelligence — you can use what proves useful to you, and ignore what isn’t useful to you. Gosh, it’s almost like they designed it to work like that.

What???!!! 

Did you see their promotion? Apple promoted Apple Intelligence for selling iPhone 16 series! 

I understand that Apple´s AI approach is challenging and different from regular LLMs as Apple focuses on local based LLM feature (That´s why Apple´s approach is harder and more challenging than any cloud-based LLMs).  

But given the size in terms of their market cap., it is still embarassing and ridiculous that they are so slow and they are dummy to release something what they have promised for 16 series. 

My bet is that Apple will never roll out a personalized Siri with such clunky codes. 
Apple has shown strange moves recently: Normally, Apple is late, but does better. Apple launches when they are ready.
But it is surprising to see that Apple hits the market with half-baked features and new products like Vision Pro (Yes.. For me, it is a half-baked item which is forgotten as of today). 

Something is rotten in the states of Cupertino. 
Maybe, they need to check if their headcounts are not enough or not capable to do their jobs? (John...Do you feel addressed??)

This is what happens when a supply chain god starts listening to marketing and not the engineers.

The engineers should have been allowed to come up with at the very least a decent beta before this was announced.

It wasn’t a selling point for me. I make sure to keep it turned off. I’m tired of everything shoving AI down people’s throats