Meanwhile, "Hey Siri, volume down." Siri: "You want me to play music at 100%?" Hey Siri, add cereal to the shopping list. Siri: "Who's speaking?" HomePods are playing music - Hey Siri, "volume down" Siri: "Turns down the volume on my nearby iPad which isn't even playing music." Hey Siri, set a timer for 15 minutes. 15 minutes later the HomePod mini alarm in the bedroom way down the hall goes off instead of the 2nd gen paired HomePods 10 feet away. My son asked me why I was in a bad mood last night at dinner. I told him because I had to yell at the idiot Siri just to get her to respond and when she does, it is stupidly wrong. Apple, fix this ridiculous situation ASAP! I'm really, really tired of Siri...and you are losing the people...wait, maybe that's a good thing!!! Ha!
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.
Apple Continues Losing AI Experts to Meta
- Thread starter MacRumors
- Start date
- Sort by reaction score
Apple AI has already done the most useful thing it can - making 16GB base Macs. Now it can go.
Typically in that scenario, you are redeployed, laid off or don't have your contract renewed. None of those were applicable here.
there goes the last 8% that still hoped apple can fix Siri
what are you talking about AI is the future that will be imbedded into robots to replace the labour force its a good return of investment
I don't know why I read this thread. I knew what a lot of the comments were going to be: Hur hur Siri bad Apple can't AI hur hur
I don't know what the Foundational Model team is supposed to do about Siri, since they're working on building foundational models, not Siri.
I don't know what the Foundational Model team is supposed to do about Siri, since they're working on building foundational models, not Siri.
In a very short timeframe, this will be true for just about every single tech job out there. Scary.
The cloud, AI, and technology is paving way for mark of the beast. Check out patent 666 owned by microsoft.
Whilst the experts are obviously, likely to have been tempted by the huge compensation, for so many to leave it also suggests a lack of faith in Apple’s AI development and problems behind the scenes. The evidence for this is Apple publicly admiring its been having problems and having to delay some of its AI products.
I don't think it's so much the pay that is causing them to jump ship as it is the damage to their resume the longer they stay. Apple has come up with a big fat goose egg with AI while seemingly much smaller companies have seemed to leverage it quite well. Hopefully they'll just break down and buy someone else's model, one that comes with a new virtual assistant to replace Siri. The money at Meta is an added bonus. I would eat brake pads for 200M.
AI is hot garbage and has been shown to reduce human cognitive abilities. People now rely on AI instead of their own brain. Quite frankly, lazy people would rather let their own brain turn to mush just so they don’t have to think for themselves. When I hear that Apple is “behind,” I have to ask, “Behind on what? Behind on having a chatbot that usurps my ability to think for myself?”
Good riddance. Apple knows that the age of giant know-it-all AI models is coming to a close. Just like humans, AI models only need to have an average level of knowledge, relying instead on tools and Internet data sources to provide up to date accurate information. The real AI experts at Apple are the ones who have been figuring out things that matter -- like the folks who are solving real problems like these: https://appleinsider.com/articles/2...m-at-ai-hallucinations-and-true-conversations .
A fourth Apple artificial intelligence expert has left the company to join Meta, reports Bloomberg. Bowen Zhang, who was on Apple's foundation models team, is the latest employee to abandon Apple for Meta.
The leader of Apple's foundation models group, Ruoming Pang, was one of the first Apple AI researchers to join Meta. Since then, several employees who worked under him have also left for Meta. Meta is aggressively hiring for its Superintelligence Labs, an AI division that's building advanced AI systems capable of performing at or beyond human-level intelligence.
Meta CEO Mark Zuckerberg has been offering AI engineers massive compensation packages to lure them from other companies, and Pang reportedly received over $200 million. Pay from Meta reportedly includes a high base salary, a signing bonus, and stock awards, and the money offered to Pang exceeds the compensation of almost all Apple employees except for executives. Presumably, the other AI engineers that left Apple have also received offers that Apple isn't willing to match.
Last month, OpenAI CEO Sam Altman said that Meta had been offering signing bonuses as high as $100 million. Meta has hired engineers and AI experts from Apple, OpenAI, and Anthropic. Bloomberg says Apple is "marginally increasing" the pay of its foundation models team, but is not paying at the level that Meta is.
With Apple losing key employees to Meta, it could continue to struggle to catch up in the AI race. Competitors like Google and Samsung have much more advanced AI features already, and this year, Apple was forced to delay promised Apple Intelligence Siri features until 2026.
Apple has been restructuring its AI teams, with AI efforts now overseen by Apple software chief Craig Federighi and Mike Rockwell, who led Apple Vision Pro development. Rumors suggest that Apple is considering using technology from Anthropic or OpenAI for future AI features, including an LLM version of Siri, rather than its own models.
Apple's discussions to rely on third-party AI technology have reportedly led to falling morale on the foundation models team that is now losing employees to Meta. Multiple engineers are reportedly actively interviewing for jobs at other AI companies, while Apple executives are aiming to reassure team members that it remains committed to in-house AI development.
Article Link: Apple Continues Losing AI Experts to Meta
It'll be interesting to see how much cash they can get out of Meta while chasing a pipe dream that LLM technology can't deliver, and government regulation will eventually prohibit.
“AI” is the most oxymoronic of all terms ever conceived by the human race.
Given how Apple is bad at AI, it is probably a good thing. Poor Meta 
The news here really is that Apple has 4 AI experts worth poaching 😂
The news here is that Meta still exists, seeing as how their social platforms are steaming piles of dog crap littered with AI-generated slop.
Honestly, I kinda want to turn off Siri too. But then I remember I wouldn't be able to set alarms or timers without saying a word... Ugh, that's a pain. That’s actually one of the few things Siri’s pretty good at.
I think this is arguably one of the key reasons behind Apple’s current struggles in AI. Apple has been overly focused on edge AI applications, following a highly closed and versioned development model. Now, to be fair, there's nothing inherently wrong with that approach. But in today’s environment, that space is more of a secondary market.
The real battleground is large-scale, cloud-based AI — which inherently demands massive compute and rapid iteration driven by fierce competition. That’s where Apple is basically absent.
So, if the so-called “experts” still argue that it’s no big deal Apple is losing talent or market share — because Apple has always been about smaller, private, controllable, on-device intelligence — then that line of thinking practically explains Apple’s failure. It’s a textbook case.
----
That said, I want to add something. Personally, I’m actually more interested in edge AI — on-device, privacy-focused intelligence. So when I say it’s a "secondary market" or refer to it as "small AI," I don’t mean it’s unimportant. What I mean is: in today’s environment, building for this space is incredibly difficult.
It's hard to imagine a 13B parameter model running locally and responding in milliseconds. Maybe someday, with better model compression and more powerful hardware, this area will rise in importance — but that time isn’t now.
To use a software development analogy: starting to heavily invest in edge AI now — or even years ago — feels like over-engineering. It’s like designing for a problem that hasn’t really arrived yet, and in doing so, pouring in nearly all of your time and energy. That approach just ends up leaving you underprepared when it actually matters — when the real competitive moment hits.
Last edited:
Sorry, I didn’t catch that.
same as Detroit Pistons has F1 drivers 😂
like being an Apple Ai expert is best job ever, like being an Olympic swimmer in a country with no swimming pool
iPhone is BlackBerry now.
Cool for a while but soon too antiquated.
How long will we keep our iPhones until we finally jump to modern OS’s?. What will be the last iPhone?. iPhone 19?.
Really, Siri is soooooooooo pathetic. There has been basically no improvements on Siri since over 10 years. Can’t even make a decent calendar appointment from a full sentence in natural language.
Images correction with AI is nightmarish, etc.
Cool for a while but soon too antiquated.
How long will we keep our iPhones until we finally jump to modern OS’s?. What will be the last iPhone?. iPhone 19?.
Really, Siri is soooooooooo pathetic. There has been basically no improvements on Siri since over 10 years. Can’t even make a decent calendar appointment from a full sentence in natural language.
Images correction with AI is nightmarish, etc.
AI is basically just a compression of human knowledge and experience — a tool that responds in an interactive and somewhat predictable way. Honestly, I think most of the controversy around AI just comes from the word “intelligence.” People get caught up in the label, but at the end of the day, it's still just a tool.
Now, about the concern that AI is making people "stop thinking" — I don’t think that’s completely wrong, but it definitely overestimates AI’s influence. What it really does is simplify how we access information. And let’s be real — humans have always looked for the most efficient, least effort path to get things done. That’s not unique to AI. We’re not so different from animals in that way — we just want more. (And hey, when it comes to building dams, I’d gladly ask a few small animals for tips!)
What I’m saying is: most successful or exceptional people tend to be highly self-disciplined. These are the kinds of people who, no matter the era, will always find ways to use every available tool to stay ahead. If AI makes it easier to be lazy, it also makes it easier for disciplined people to achieve even more, even faster.
As for the ones who’ve always relied on shortcuts and instant answers? They’d be that way with or without AI. Let’s face it — for a lot of people, their brain’s just a decorative organ. Once you realize that, it’s clear: AI isn’t making things worse. It’s just making the gap more obvious — and more helpful — for those who are ready to take advantage of it.