Gurman’s column today is a pretty big indictment of Apple’s AI plans. He starts by talking about how the current AI features released so far are underwhelming and half-baked. Like the ChatGPT integration which is not smooth or nearly as useful as just using ChatGPT by itself. The genmoji are like a toy compared to the advanced features of ChatGPT and Claude (which Macsparky recently wrote about here).
Welcome to Apple’s AI crisis.
He then talks about how Siri has two brains, the old Siri that sets timers, etc. And the new supposedly improved Siri that is due to come out in May.
That’s why people within Apple’s AI division now believe that a true modernized, conversational version of Siri won’t reach consumers until iOS 20 at best in 2027.
That would mean Apple is a half-decade late to the game — an even bleaker timeline than many of us imagined."
He concludes saying that changes need to be made at Apple for them to have a chance of fixing this.
That’s left Apple at a make-or-break point. Clearly, the company isn’t moving fast enough internally to create the underlying AI technology it needs to keep up with the competition — and that suggests a change is required."
If I had a pound for every time Apple’s been pronounced doomed because it’s “late” into the new new thing, I might not be rich, but I’d certainly have quite a few pounds.
It’s pretty clear Apple is getting it’s AI clock cleaned. Apple Intelligence is such a disappointment. I almost never use it. I think Apple will eventually get its act together, but it really seems to be behind.
IMO, that’s true. As I recall Gurman said on TV a few weeks ago the real “new” Siri wasn’t expected until Spring 2026. I really don’t care because I mainly use Siri to launch apps and control the volume of podcasts.
Apple’s not doomed because the world of AI is racing ahead while they are trying to get their horse in the starting gate. IMO, the only thing that could hurt Apple is if the market stops seeing them as a growth company.
I would like to see a new Siri that could integrate with the 3rd party iOS/iPadOS apps that I rely on. And how Apple plans to work with the major AI’s that businesses are adopting. As long as Apple’s devices can continue to work with the software that “runs the world” they will likely stay at or near the top.
I don’t think Apple is doomed. No matter which company wins the AI battle, people will still need computers and smartphones to run the AI apps. It’s just sad that Apple really had a head start with Siri and then never really worked to advance it. And even now with 2+ years of LLMs they aren’t even close to an improved Siri. This is what I really have a hard time understanding.
I am not a fan of Siri and tend to use it more on the watch than anywhere else. Apple usually gets there in the end. I can use Claude or ChatGPT to do what I need. I loose nothing by waiting for Apple to develop a local or more secure AI that can integrate email, contacts etc with the whole ecosystem. Is there some competitor that is doing this better ?
I don’t think they’re “doomed” as a company - as others have said, the making of great hardware will still be valuable.
It’s also hard to imagine them ever catching up at this point. The most recent models (I use Grok primarily) are so good. I’m use AI/GPTs in my job every day and I’ll bet Grok saved me 4-8 hours of work last week AND made the quality of the work I did better. It’s a tool, not a replacement for my abilities, but it’s a darn good tool.
In the areas where Apple has released a feature that even tries to be this useful, they appear to be 2-3 generations of GPT behind where everyone else is. The image generation is garbage compared to their competitors. I’d say even the image “improvements” like erase background can’t hold a candle, either. The writing assistant is at least that far behind.
On the AI front, my opinion is that they either need to buy a MAJOR player in this space or they need to give up their own features and work hard on integrating well with all the big players. Apple’s sweet spot has always been integration - usually of their own services and not other players - but I think in this space they’re going to have to learn to play well with others and be the aggregator of the best capabilities in the market. They’re wasting money development and releasing the embarrassing “Apple Intelligence” features I’ve used so far.
I was thinking that too. It was widely rumored that the car project was focusing on an autonomous car, I would think some of that work would overlap with their AI efforts or is it too specialized for navigating a car to be useful in the chat LLM area?
Would the government allow the most valuable company in the world buy another major company?
My guess is they will continue to work with the big players regardless of what they do on their own hardware. IMO, the Mac would have never become a general purpose business computer if Microsoft had not continued to make MS Office for Mac. And Apple didn’t get Mail.app working with MS Exchange?
As it is, both Microsoft and Google have already rolled out their own AI’s to their business customers so it will be interesting to see what on-device Apple Intelligence can bring to the party.
I use ChatGPT, DeepSeek, Papago, and others but sparingly. It is a conscious choice when I do. Tight integrations with iOS are not something I want nor likely to use. So no, I’m not bothered that Apple might have been late to market and likely to be even later with something fully-functioned (based on someone’s artbitrary definition).
Android and Windows have a competitive advantage in this field. They have less qualms with apps talking to each other so they can give LLMs system wide capabilities that Apple has historically been loathe to let iOS and iPadOS (and to a lesser extent macOS) have. Apple’s historic strength is its greatest disadvantage here, the walled garden. Combined with Microsoft and Google’s cloud and web strength they have a lot of catching up to do.
I would prefer they leverage their hardware advantage to make Apple a competitor in the Cloud space and web space and leave the chatbots to the others so Apple can build it out privately and build a real private alternative. But alas that is not likely to happen.
Siri is a butler or a secretary. No more. I might hope it stays this way. ps … Anyone remember when you could give voice commands to the mac "Computer … "???
How far does one really want the macOS to expand on this front to become “everything, including the kitchen sink”? Should the next vision after AI be that we must be able to watch VR movies native in macOS as an inherent, built-in capability because … well … gosh darn … being able to see your pets or grandkids in 3D motion is what everyone needs as the next hottest thing.
I hope that Apple will leverage the ability to allow developers to build effective, efficient abilities to go out to and bring back information from external (Web-based) or third party AI apps. I sincerely hope that they don’t re-invent and then have to continue to sustain a “better” AI wheel. I sincerely hope that they also do not pin everything new and sparkling about AI (and the kitchen sink too) firmly into the existing macOS operating system. Othewise, we may someday be talking about the macOS that tried to interpret our needs with AI but was never developed any further after the first go around.
I’m surprised how little the Apple fans are willing to concede on this one. Grok is not tied in to my os, neither is photoshop. They are generations of technology ahead of anything Apple is doing. It’s not about protecting my privacy, nor is it about being too early to market. Apple simply has not, in any way, demonstrated they have even a fraction of the technological capabilities of the other players in this space. And it is not because they aren’t trying. Siri is bad. Image generation is bad. Writing tools are bad. Email summaries are bad.
I love Apple hardware but I stand by what I said above and am surprised how many excuses are being made for Apple in this thread. Is it not okay to just admit they whiffed on this one and need a different strategy?
In what might once have been called top posting but now is called follow on, I add a specific example.
Consider go be an existing or to be developed AI tool as a Web-based or self-standing app. What I do not want versus what I would enjoy.
NO: Siri, summarize the selected articles in a Word document.
YES: Siri, tell to summarize the selected articles in a Word document.
YES: Siri, summarize the selected articles in a Word document using .
I agree. When your neighbors are all owning custom-built 9 speed blenders that you can borrow and you only have a rolling pin, it is best to realize where the smartest investment of your efforts should be to give your guests the best frozen margaritas for the upcoming party. It is not in trying to convert the rolling pin into a 10 speed blender.
At this point, the AI “revolution” looks very much like a bubble. Vast investment when none of it is making any profit at all (yet) and it’s not clear what needs to be done to do so. There’s no development path for quite a few of the things that might be needed: people are seeking a breakthrough that hasn’t arrived yet.
really powerful on-device AI will be more or less invisible to the user IMHO. It’s things like dictation “just working”, predictive text being accurate and more than the next word, it’s automatically naming people in your photos etc. Apple already has a LOT of that kind of technology
the “neural engine” in Apple Silicon is miles ahead of the competition
Apple may be wise not to hurry. It’s possible they may have made a ghastly strategic error and will never catch up, but I’m far from convinced that AI will look like the son of Claude or Chat-GPT in a few years anyway.
go be an existing or to be developed AI tool as a Web-based or self-standing app. What I do not want versus what I would enjoy.
It is not in trying to convert the rolling pin into a 10 speed blender.