Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Vision Pro Creator Taking Over Siri After Apple Intelligence Setbacks
- Thread starter MacRumors
- Start date
- Sort by reaction score
Vision Pro may have been a flop but that’s what happens when you bring a product that no one asked for to market. Technologically and engineering wise, the Vision Pro team delivered, something the Siri team under Giannandrea has been unable to do for over a decade.
Since the VP was such a massive success, they put its creator in charge of Siri, makes sense...
Likewise for the wretched person responsible for making macOS look like iOS
It's not a commercial success, but it largely does what it set out to do. And more importantly, it's relatively bug free and - if you ignore that it's a solution in search of a problem - well liked by people during the short time that they use it.
Hey Siri, where I can find a nearby pharmacy?
Siri: I don't know, but you should buy the Apple Vision Pro!
Siri: I don't know, but you should buy the Apple Vision Pro!
Oh my. Vision Pro is developing well. Hope it doesn’t impede the AVP development.
Siri is dead in the water and useless to be honest.
Siri is dead in the water and useless to be honest.
So they’re promoting the guy who heads up a product no one wants to head up Siri, a product everyone hates….
I think it’s time for the head of Apple to change!
I think it’s time for the head of Apple to change!
"I've found a website that can help. I've sent the results to your Apple Vision Pro. Oh, wait? You don't have one? Would you like me to purchase one?"
The fact that this leaked to Gurman, before even the employees know, means that Cook has top executives who are not loyal to Apple whatsoever.
They would rather leak info and force their employees find out about these changes via Gurman, than stay silent and loyal to their workforce and Apple’s ethos of secrecy. Not good.
And does Gurman pay his sources? If he does, there is no way he pays what these executives make at Apple. So then, what is their incentive to leak this kind of info besides causing chaos?
They would rather leak info and force their employees find out about these changes via Gurman, than stay silent and loyal to their workforce and Apple’s ethos of secrecy. Not good.
And does Gurman pay his sources? If he does, there is no way he pays what these executives make at Apple. So then, what is their incentive to leak this kind of info besides causing chaos?
I don't think we'll be able to call Vision Pro a "flop" until we hit Year 5. Apple Watch was large in part a flop for the first several years before cheaper models drove sales.
This is hilarious with the news today that the lead for Vision Pro has now taken over Siri 2.
Apple's biggest product flop/worst selling product in recent history is now the lead for Apple's second biggest product flop.
Why not bring in the lead engineers for AirPower?
Apple's biggest product flop/worst selling product in recent history is now the lead for Apple's second biggest product flop.
Why not bring in the lead engineers for AirPower?
People actually want and use wearables like smart watches
Mainstream general users don't like putting on scuba masks that isolate them from their surroundings
Chief Floperator is now moved over to an even larger Flop division
This. The Siri AI flop is just the latest of Apple’s mismanagements, there’s no longer anyone who cares about the ‘little things’ that made software that just works, only about maximizing profit, squeezing the last cent out of every customers, make them climb up the ladder in hardware, force them into services and so on.
And I get as a publicly traded company there’s no way around it, but the CFO should handle that (and tone it down), the point is that there’s no leader at Apple.
On the other hand, the biggest issue with Apple Intelligence is that it barely works and half of it hasn’t shipped.
Everything advertised about the Vision Pro works
VisionOS is rock solid.
Vision Pro actually shipped, and not only did it ship, it’s shipped pretty much right on schedule. Apple said “early 2024” and the thing was out 33 days into the new year.
Like the product or not, the point is that Mike gets results.
This isn't a John Giannandrea issue. This is a Craig Federighi issue. At an executive level this should have been managed and adjustments made long ago - software across the board may destroy Apple's supremacy (if not already). Leadership in that department needs reimagined completely. Hardware is crushing it. Software is not.
I work on an AI based product at my company. It's more in the role of assisting the user with automated tasks than in raw content creation, so it's mildly analagous to Apple Intelligence.
I often describe the LLM as "squishy" or liken it to a recalcitrant teenager who misunderstands what was asked, smarts off, or refuses to do what's asked. It's not deterministic like a piece of Python or Swift code. It's ...squishy... like trying to talk to a person.
It is really easy to put together a demo that works well, especially when people unconciously prompt in a way the AI understands. Then, when you release to the real world, and people use different phrasing or aren't specific about what they want, you get strange results. It's like the LLM tries to give a pleasing or placating answer, and if there's not enough to go on...that's when it hallucinates.
I deal with this all the time. It can take a lot of system prompt engineering to get something that delivers results to a wider audience. The effort to get all the prompting, and maybe even fine-tuning the model, correct can be a huge time sink compared to the coding.
Given that, I'm not surprised that Apple had a hiccup between demo and execution. And, I think they'll figure it out.
On the other hand, I'm an avid, daily Vision Pro user. Not all success is measurable by financial success. The Vision Pro is a technological marvel, and I think, a bit ahead of its time. More than a year later, I'm still in awe of the thing, realize it's got flaws, enjoying every improvement. Makes me feel like the Macintosh in 1984 again. Paul Meade taking over leadership on Vision Pro is a hint that development will go on.
I think we're too used to Apple hitting every ball out of the park. This isn't a failing company. This is a company that's facing reality. Not everything is a blockbuster. Apple is having its Cars 2 moment, and...look, Pixar is still making great movies, and a bad Pixar movie is still pretty darn good.
I often describe the LLM as "squishy" or liken it to a recalcitrant teenager who misunderstands what was asked, smarts off, or refuses to do what's asked. It's not deterministic like a piece of Python or Swift code. It's ...squishy... like trying to talk to a person.
It is really easy to put together a demo that works well, especially when people unconciously prompt in a way the AI understands. Then, when you release to the real world, and people use different phrasing or aren't specific about what they want, you get strange results. It's like the LLM tries to give a pleasing or placating answer, and if there's not enough to go on...that's when it hallucinates.
I deal with this all the time. It can take a lot of system prompt engineering to get something that delivers results to a wider audience. The effort to get all the prompting, and maybe even fine-tuning the model, correct can be a huge time sink compared to the coding.
Given that, I'm not surprised that Apple had a hiccup between demo and execution. And, I think they'll figure it out.
On the other hand, I'm an avid, daily Vision Pro user. Not all success is measurable by financial success. The Vision Pro is a technological marvel, and I think, a bit ahead of its time. More than a year later, I'm still in awe of the thing, realize it's got flaws, enjoying every improvement. Makes me feel like the Macintosh in 1984 again. Paul Meade taking over leadership on Vision Pro is a hint that development will go on.
I think we're too used to Apple hitting every ball out of the park. This isn't a failing company. This is a company that's facing reality. Not everything is a blockbuster. Apple is having its Cars 2 moment, and...look, Pixar is still making great movies, and a bad Pixar movie is still pretty darn good.
don't worry guys, Lord Bean Counter knows what he's doing.
So what's going to be the iPhone 17 Pro Killer Feature that forces people to upgrade?
The iPhone 16 Pro was all about AI.. and they did not deliver any AI on the 16.. so the 16Pro better get the same AI features the 17Pro will.
Anyone that got the 16Pro should be let down and furious. These features were advertised during the NFL games all fall... nothing is here besides a crappy playground where I can make emojis.. that's not even AI..
The iPhone 16 Pro was all about AI.. and they did not deliver any AI on the 16.. so the 16Pro better get the same AI features the 17Pro will.
Anyone that got the 16Pro should be let down and furious. These features were advertised during the NFL games all fall... nothing is here besides a crappy playground where I can make emojis.. that's not even AI..
For those of us who actually own the Vision Pro it’s a phenomenal success. The hardware is great. The software is great and it’s getting better.
For those who don’t wanna drop $5000 I guess I understand their view that if something is not a mass market success then it must be a failure…
I'm wondering if Giannandra made a lot of enemies inside Apple. I think Gruber's right in that this is a turf war Craig just won (as Rockwell now reports to him), but maybe he (Giannandra) pissed off a lot of other people.