John Gruber Says 'Something is Rotten' at Apple

[HouseLannister]

While I get that people are genuinely upset and feel lied to, I haven’t seen one story of someone trying to return a phone they bought six months ago because it does not have the advertised features. Where are the switcher stories where they fled to Android? I would think the media would be all over this stuff. Are people really mad or just disappointed and sad?

 

[klasma]

I'm guessing part of the problem is that Apple has been expecting to use the LLM approach to getting Siri to reliably do the things that the original Siri company (that Apple bought) demonstrated 15 years ago, along with the additions that Apple wants to make to it to turn it into a modern personal digital assistant (on-device awareness of the user's data, intent, contents of the screen, images of the user's immediate surroundings, etc.), but as everyone has learned, LLMs currently can't do those things properly consistently enough to meet Apple's (and our) goals for them.

The issues with LLMs were already crystal clear a year ago if you worked with LLMs, with nothing on the horizon to really solve that reliability problem. As Gruber says, the technically responsible people at Apple can’t possibly not have known this. And yet they made promises based on wishful thinking.

 

[wanha]

They put that PM person in charge of Siri a month or 2 ago. What has been transpiring over the past couple weeks is the outcome of the work of that person, they have known about it for a while, hence the change back then.
Grubers “PhD knowledge of Apple” (that is his claim) should have highlighted this months ago, he’s just riding the wave now.

Not sure what more you want, friend. Gruber says in his piece very explicitly that he should have known better and details all of the instances and signs that he missed but clearly shouldn't have.

IMO suggesting this piece is "just riding the wave" is incredibly naive - it's a scathing public criticism of Apple's integrity from one of its biggest, longterm supporters

 

[Felix Lyran]

Hopefully not a controversial take, but in my view it would have been better for Apple, credibility-wise, if they simply ignored the AI bandwagon and just gradually implemented these new features. Whether consumers think of new and emerging features as "AI" or not doesn't really matter, so long as the features are there, they work, and they allow users to do actually useful things.

I know Apple was previously criticised by some commenters for not doing enough to "respond" to the AI trend, but I actually think not responding was a good idea, because so much of it is just junk, a glorified version of copy-and-paste, or, at the worst end, a "hallucinating" plagiarism machine. (I'm not necessarily talking about Apple Intelligence specifically here, but about "AI" broadly.)

Because so many companies have rushed to get in on this and to push it out to consumers—and "push" is apt, given how many apps are pushy about having you use their latest AI-assisted whatever—it might actually be better, credibility-wise again, to have a few companies take it slower, focus more on usefulness, and not foist it upon users in a way that just stops short of implying that it'll change their lives and, oh, here's our new AI-assisted thing, would you like to use it to generate a...

 

[Belisama]

John Gruber should get a life...
We are in the "post truth" era, the century of "alternative facts".

We are flooded with lies every single second, from the bottom of the streets to the top of our institutions.
Deal with it! We get what we deserve. What we chose.

Who is stupid enough to believe Apple, Meta, Google and so many other BS? Who is stupid enough to believe that we are about to colonize Mars? Who is stupid enough to believe Elizabeth Holmes, Sam Bankman-Fried, Musk, Cook, Zuckerberg and hundreds of other corrupted CEOs of companies that know exactly what to do to s***w us ?
I know... potentially everyone is stupid enough.

So deal with it !

 

[JosephAW]

Apple spending huge amount of resources, time and money to develop ai and then a large amount of their user base immediately disable it on their devices. We’ve gone thru several updates where Apple reenables ai after an update just to have users turn it off again. I’m sure they can see this back at Apple. The next step will be there is no “off” switch.
And they just released an iPad that can’t even run ai after the rollout.

 

[Kiwamu]

Hopefully not a controversial take, but in my view it would have been better for Apple, credibility-wise, if they simply ignored the AI bandwagon and just gradually implemented these new features. Whether consumers think of new and emerging features as "AI" or not doesn't really matter, so long as the features are there, they work, and they allow users to do actually useful things.

I know Apple was previously criticised by some commenters for not doing enough to "respond" to the AI trend, but I actually think not responding was a good idea, because so much of it is just junk, a glorified version of copy-and-paste, or, at the worst end, a "hallucinating" plagiarism machine. (I'm not necessarily talking about Apple Intelligence specifically here, but about "AI" broadly.)

Because so many companies have rushed to get in on this and to push it out to consumers—and "push" is apt, given how many apps are pushy about having you use their latest AI-assisted whatever—it might actually be better, credibility-wise again, to have a few companies take it slower, focus more on usefulness, and not foist it upon users in a way that just stops short of implying that it'll change their lives and, oh, here's our new AI-assisted thing, would you like to use it to generate a...

.... on the other hand, shareholders think this is a gold mine and want Apple to go all in.

 

[Bogstandard]

Internecine warfare in the boardroom regarding succession...me thinks.
They are likely Too distracted to bother with the rest of the company.

 

[bondr006]

Glad I decided not to give up my Alpine Green 13PM. It is the best iPhone I've ever had. Don't give a darn about Apple AI. If I want to use AI, there are already a ton of apps for that. Even though she has improved somewhat since her birth, I do wish Siri was smarter, as I do use her regularly. For the most part, I am ok with Apple. I just wish they would work more on perfecting the features that matter the most, stop trying to impress us with smoke, vapor, shock and awe, and just be transparent. I am mostly interested in the hardware improvements. I really miss Steve Jobs and I agree that Tim Apple needs to go. I really think that Jony Ive would have made a much better CEO than the bean counter.

 

[alchemistmuffin]

This obviously isn't the first time that Apple has failed to deliver. However, Gruber said other examples like the canceled AirPower charging mat "tended to be around the edges," whereas he believes that generative AI is going to be "big" and "important."

It's not the delay by itself that bothers Gruber. He said the true "fiasco" here is that Apple "pitched a story" last year "that wasn't true":

Gruber said the personalized Siri features announced during the WWDC keynote last year were merely conceptual, and therefore "$&@!%?¥”

This exact fiasco is one of reason why Apple cancelled the spring keynote and opted to release everything via press release. The only part Apple filmed for the keynote was the iPhone 16e. In fact, the highlight of the keynote was to be the M3 Ultra and Mac Studio but that portion was never filmed because Apple cancelled the event by then.

Apple could be in trouble big time because they were way over-optimistic on Apple Intelligence timeline and have everything done on device that they announced things way ahead of schedule thinking they can deliver before this year’s WWDC. Obviously they hit the wall, because of hardware limitation. They could compromise and go via cloud AI method but that is something Apple won’t do, not just privacy but attention to detail.

 

[kiranmk2]

I'm not sure I agree with this article and the large pile on. Apple does do a lot of annoying and consumer unfriendly things and has had several technology fails, but I don't think this puts them in a bad light compared to other CE companies.

Regarding AI/LLMs specifically, it seems fairly clear that of the big tech companies, only Google was taking LLMs seriously with their own in-house development. Then came the release of ChatGPT 3 in late 2022 (and the associated MS CoPilot). This launch made the large companies (including, not only Apple & Amazon) suddenly take notice that their (initially market leading) AI solutions had fallen far behind. It even made Google, who had been developing Bard, rush it to market before it was ready to signal to investors that they could keep up. In fact, it was reported in the WSJ (described in this AppleInsider post) that Craig F only started to understand how far behind they were when he used CoPilot around Christmas 2022 and that before this, he didn't really get the importance of LLMs, preferring to focus AI on image/video-based features.
With all the hype around "AI" (meaning LLMs) and the real threat they could pose to the incumbants (e.g. an LLM smart speaker could overtake Alexa/Google assistant devices to be, LLMs could replace Google's search engine (and main revenue stream) everyone and their dog fell over themselves to communicate to investors and the market that they wer bringing AI to their products and wouldn't be left behind.

Regarding Apple Intelligence and the iPhone 16 range, again the fact that the iPhone 15 (non-pro) came with an A16 chip/6 GB and did not support AI is probably related to the fact that it was planned out in 2022 before the ChatGPT moment and the thought that lots of RAM would be needed for AI models within 1-2 years wasn't even a consideration. As above, I think that Apple had to show the market that an LLM was coming to the Apple ecosystem - hence the WWDC annoucement. The fact that complex queries have to go to ChatGPT again shows how far behind Apple are, but getting something to market quickly was key. I imagine that the plan to drip feed out Apple Intelligence features over the year probably seemed like a good idea (it's happened many times previously that WWDC-announced features are only added into iOS .x releases) and to be honest I believe that Apple acted in good faith with the WWDC "demonstration." What has gone wrong is that, as Apple and Amazon have found, it's difficult to implement an LLM system from scratch into an existing product. Amazon also planned to add their own LLM to Alexa, then had to go to Anthropic for the LLM and still struggled to get it working properly - we'll see how it performs on release next quarter). Now, questions around whether they were naive to think everything could be done within 12 months, whether they should have officially delayed the introduction date sooner and whether they should have built the entire iPhone 16 advertising campaign around AI are all valid and should be reviewed internally.
For me, this goes back to the common advice of: don't buy a product based on promises of what it will be able to do in the future; buy it for what it can do today. I typically buy refurbished iPhones from the previous generations because I often see big new features not really maturing for 6-9 months (off the top of my head, TouchID on the iPhone 5S couldn't do much apart from unlocking the phone and authourising App Store purchases until Appe developers started supporting it and Apple Pay on the iPhone 6 didn't come to the UK for months after launch).

Regarding Apple under Tim Cook, I think a lot of people either don't remember or didn't really experience Apple under Steve Jobs. Even back in 2009 when I first got a MacBook, it came with too little RAM and storage and the upgrade prices were a rip-off (although you could upgrade them yourself back then). iPhone & iPad storage tiers were still too low & expensive to upgrade and products shipped with defects (nVidia GPUs). So, I don't see things that much worse under Tim Cook. I think the range of products that have been heavily reported to be under development (Car, non-invasive blood glucose sensors, AR glasses etc) as well as Apple Silicon show that Tim is willing to open the purse strings to go after new products, but again questions around why some of these large investments fail are very valid - look at Xiaomi: a similar CE producer that backed development of a car and are now selling loads of them at prices from $30k upto $100k in China.

TLR - Apple has many failings, but I don't think the Apple Intelligence failure is as bad as is being made out (and buy a product based on what it can do today, not what it may do tomorrow)

 

[sunapple]

While I get that people are genuinely upset and feel lied to, I haven’t seen one story of someone trying to return a phone they bought six months ago because it does not have the advertised features. Where are the switcher stories where they fled to Android? I would think the media would be all over this stuff. Are people really mad or just disappointed and sad?

I think most people are still unaware of the situation, it could gain traction in the near future.

 

[amaze1499]

If Mr Gruber speculates it must be true?

What an inside… every company struggles at some point. What’s new?

 

[SBlue1]

Oh, he mad.

Let's face it, Apple dropped the ball on AI and is a step behind the market. They've gotten so lazy that it is hard to be surprised anymore they are mailing it in. The Dynamic Island was really the last jaw dropping software moment and that is only because of the false, silly leaks about the gap.

In my personal opinion AI and personal assistants will be huge and game changers in the future. AI is a huge help and a giant leap in productivity in certain areas in my business and is really helpful for me personally. Seeing what is possible and knowing I can only set alarms on my HomePod mini is very frustrating. A dumb Siri and Apple Intelligence lagging behind every other competitor can be Apples Nokia moment and can bring the entire company down from mainstream back to being niche.

 

[HouseLannister]

In fact, the highlight of the keynote was to be the M3 Ultra and Mac Studio

Haha. No. Apple knows that not having an M4 Ultra Studio and Mac Pro ready by Spring was not going to be something you make a keynote for. Apple can’t even get every computer on the same chipset each year and that doesn’t need extra attention. That was always going to be a quiet launch.

 

[kamyk35]

I use Gemini flash all the time on my S25 ultra- it's deadly accurate but I think the early mistakes tarnished the reputation.

Gemini is less than 40% accurate with general knowledge questions (according recent tests), just behind Chat GPT, so it's certainly not deadly accurate, also the S25 Ultra is just hype and certainly not deadly accurate either 🤣

 

[sunapple]

I imagine that the plan to drip feed out Apple Intelligence features over the year probably seemed like a good idea (it's happened many times previously that WWDC-announced features are only added into iOS .x releases) and to be honest I believe that Apple acted in good faith with the WWDC "demonstration."

Phased roll out of features is fine and not the issue in question. All AI features that we have today were shown in on-stage demos at WWDC24. Smart Siri that's been delayed by a year was not in any live demo, it was only mentioned in video and ads. They might not have had it working at all at the time. So the argument of the blog post is that this was shown prematurely on WWDC and definitely used in marketing way too soon. Huge gamble. Tim Cook should have ran a tighter ship.

Now, questions around whether they were naive to think everything could be done within 12 months, whether they should have officially delayed the introduction date sooner and whether they should have built the entire iPhone 16 advertising campaign around AI are all valid and should be reviewed internally.
For me, this goes back to the common advice of: don't buy a product based on promises of what it will be able to do in the future; buy it for what it can do today.

While I agree, Apple does/did have a lot of customer trust on that front. Their track record is pretty great compared to other companies. This makes the current situation so curious.

 

[sigmasirrus]

Totally agree with Gruber. They mislead the public into believing they’d get some great AI features a few months down the road - but only if they upgraded their iPhones. Analysts went as far as calling for an iPhone upgrade supercycle due to these great AI features. Most prominent/exciting being a Siri that knows about you. And that is the exact feature that turns out to be vaporware - for the entire iPhone 16 cycle! So the trust is broken. You can no longer believe what Apple tells you.

Honestly that feature would be the only useful thing I would want from AI. And Apple is the only one I trust to deliver it. Anyone else would need to put all your info in their cloud and probably data mine it for ads. But to be able to ask Siri “umm someone sent me a link maybe like a month ago to some restaurant. What was it again?” And have it find the text message would be amazing! And it’s possible. ChatGPT can answer vague questions like that. But can Apple pull it off? ……..

 

[ttyRazor]

Yes, quite misleading. They seem so far behind everyone else. They tried too hard with smoke and mirrors to convince everyone they were catching up, but now it seems they’re falling flat on their face.

I think their bigger problem is that they’re chasing everyone else’s smoke and mirrors. While they never announced it publicly, the billions they sunk into their car project was spurred by Musk’s hype of full self driving being just around the corner. They found out the hard way that the reality would be much harder and more limited, to the point of not being worth it.

While AI will find its use cases, we still don’t have a good grasp of its limits. We’re still at the point where a bunch of things seem to be in reach with enough hardware, data and training, so reading all your texts and emails might sound like something AI could get something useful from until you consider all the poorly formatted garbage that fills our inboxes, and that’s before we get maliciously crafted messages designed to trick Siri into thinking they’re relevant or actionable.

 

[iamdamian]

I personally feel he's spot on.

Steve made bold promises and made them happen with the assistance of Johnny Ive et al.

Tim is making promises he doesn't have the ability to make his staff come true.

So many balls dropped, not just this.

I'm aware there were earlier nails for other people. But not moving the charging port on the USB-C was the face falm that did it for me and Apple. The power button on the bottom of a desktop computer was the next, even bugger face palm for me. I don't care if it's used once or ten times a month. You don't put a power button on the under side of a machine. Siri is embarrassingly behind other smart assistants and has been for over a decade. Let alone this latest Ai delay debacle for Apple.

Yes you can argue Tim has done a good job, and their stock value has risen and they're the most profitable company in the world. but what does that mean if your customers a fast disengaging and see you as a joke of a company that comes out with some of the most obvious design mistakes I've seen in the tech industry.

I agree with others, it was good while it lasted but Tim DOES need to go. My worry is, is there anyone else remotely alike for better to take his place - I personally think not.

This is just my opinion and forms the basis of my long terms (decades long) relationship with Apple. I'm so close to being done with them.

 

[dpdpdp]

Daring Fireball's John Gruber today shared some strongly-worded comments about Apple's delayed personalized Siri features. Gruber is a well-known Apple pundit who has been writing about the company for more than two decades.

In a blog post titled "Something Is Rotten in the State of Cupertino," Gruber said Apple's credibility has been "damaged" by the delay:This obviously isn't the first time that Apple has failed to deliver. However, Gruber said other examples like the canceled AirPower charging mat "tended to be around the edges," whereas he believes that generative AI is going to be "big" and "important."

It's not the delay by itself that bothers Gruber. He said the true "fiasco" here is that Apple "pitched a story" last year "that wasn't true":Gruber said the personalized Siri features announced during the WWDC keynote last year were merely conceptual, and therefore "********":He was even more explicit here:Gruber said Apple's repeated unwillingness or inability to demo the personalized Siri features in action since WWDC last year "should have set off blinding red flashing lights and deafening klaxon alarms" in his head that something was wrong.

Gruber went as far as saying that Apple's culture of excellence could be at risk if this situation is not handled correctly within the company:The full post is worth a read.

Article Link: John Gruber Says 'Something is Rotten' at Apple

An Apple tradition, in principle at least (ignoring the well-known errors and exceptions), is not releasing a product until it is basically perfect. Someone in Apple bought the hype about LLMs as the long-awaited arrival of genuine AI, and this sent Apple down the same road so many other companies were traveling — premature announcements and all. But somewhere in Apple, they saw the truth: that this is not "artifical intelligence" at all, but a parlor trick of stochastic parrotting. This is a technology that cannot distinguish true answers from merely probable answers, that is impressive on common tasks but produces crazy errors when stress-tested even just a little bit. It's not just a matter of occasional "hallucinations", it's the very nature of the thing that it has no notion of reality or truth. That's a far cry from "perfect" as a tool for actual Apple users, and potentially dangerous too.

My guess: Apple is too smart to join the chorus on so-called "generative AI" at this point, and never will. What they will do instead, I don't know.

 

[Hopefully Smarter]

a truly big error. And possibly the most cringe ad campaign ever

 

[akbarali.ch]

Have you used one of the "competing" AI's? (I put competing in quotes because Siri doesn't belong in the same category.) I personally use Grok3, and it is incredibly useful when I am trying to gather information. It is like talking to the Star Trek computer, and it explains the reason for its suggestions. It is really what Siri should have been 5 years ago if Apple had bothered to invest resources into it. Isn't it just dumbfounding that Apple sitting on that pile of cash doesn't have the resources (or it is poor management?) to invest in and continue development on every product at the same time. It is always hoping from one update to another. "Oh, it might be time to start developing a new Studio Display after 4 years." Given the pace of advancement of AI, Apple will not be able to catch up. Apple is still planning its AI server farm.

Agree, chatgpt, grok, deepai, deepseekai, claude they are amazing and Siri is waaaaaay behind, infact i dont see much change from its first iteration from iphone4S frankly speaking and they never worked in that direction. I never saw any kind of signitficant update. But i'm pretty optmistic they will catch up with AI, they got hit hard in the guts with advent of chatgpt.

 

[sigmasirrus]

Absolutely rotten. You know it’s just so rotten that the rotten pieces are rotting. It’s like cheese with mold that has cheese in the mold. As Tim Cook would say I will make the best rot ever. I’ll even attach a chart showing how bad this rot is. This is a five-year graph of the rot and it’s absolutely horrendous.

View attachment 2491617

It shows exactly who they are trying to please: the shareholders, not the customer. Good for the short term, not the long term. Remember blackberry and Nokia!

 

[CheffyDude]

It seems that these days that being genuine and truthful isn't really all that important if you are rich and successful. Apple doesn't care if they mislead their customer base. They needed to sell product. They are laughing all the way to the bank.