A new report suggests that Apple CEO Tim Cook has lost faith in his AI/ML chief John Giannandrea and is shifting the robotics team to hardware headed by John Ternus.
Apple has been rearranging some of its teams after making a publicly embarrassing move to delay contextual Apple Intelligence features. The Siri team moved to be placed under Mike Rockwell, the Apple Vision Pro chief.
It seems that wasn't the only internal move planned by Apple, as Bloombergreports the robotics team is moving out from under the AI/ML organization. It will instead be under the hardware division, which is led by John Ternus.
These moves haven't changed John Giannandrea's position as SVP of AI and ML at Apple, but it indicates Apple CEO Tim Cook is taking a different strategy with the refocus. It seems Giannandrea's team will build the underlying models that will run future technologies, including the robots and Siri, but work independently of the other teams.
However, rumors from the sources providing these details suggest Cook could be planning a total break-up of the AI and ML team. It could mean Giannandrea being assigned to a new position or leaving the company altogether.
Rethinking AI
The motivations of these moves aren't known and can only be guessed at beyond the obvious catalyst of the embarrassing AI delays. Cook could be trying to ensure Giannandrea and his team aren't spread too thin while ensuring robotics hardware is prioritized and Siri is developed independently.
On the other hand, the AI and ML team could be seen as redundant as Apple rethinks Apple Intelligence and its role at the company. Rather than treating AI as a separate product line, it could be being seen as an underlying framework.
Think app versus operating system.
For example, the robotic arm will rely on AI for specific kinds of interactions and data, but the robot itself isn't AI hardware. The same goes for Siri, which would call out to AI tools for answers and summaries, but itself isn't an LLM, not yet anyway.
Whether that approach requires a specific team dedicated to building the underlying models, or if each department can build their own models, remains to be seen. Time will reveal Apple's strategy as more moves are made public and whether Giannandrea sticks around.
7 Comments
My guess is that Giannandrea will eventually find himself in a corner office at Infinite Loop (not Campus 1) with the title VP Special Projects with a handful of others below. At this time, he might find himself thinking that a change of scenery would be better for all. He has changed companies before, it won't be the first.
In any case, it appears that the previous trajectory wasn't working. It's a business, sometimes these changes must be made.
Every day I'm more convinced that Apple is spread way too thin.
The modem project was colossal in many ways. It's a moving target and the homegrown Wi-Fi chip is also rumoured. Another moving target.
Tied to that there are the 6G research efforts.
The car project will have also used its fair share of resources.
At this point the claims that Apple wasn't behind on AI are leaking more and more by the day.
Apple hardware for training/inference is also supposedly being designed although currently they are using equipment from other vendors.
The only apparent area of roadmaps following their course in relative calm is in the chip design unit and while there has been upheaval there, it has been in the completely normal range.
Executive shakeups aren't uncommon either but the amount of rumoured seat changes of late is concerning.
On the subject of robotics, perhaps there is a parallel to be drawn with the AI situation. Competitors are mass producing robotic solutions now and upgrading them at lightning speed while the most we hear about from the Apple rumour mill is a robotic arm on a unit of some kind.
It's anybody's guess as to what the real state of Siri is and I've been wondering about the underpinnings of Apple's mobile OS offerings for a few years now. I feel sure that some major re-working will be announced at some point to pave the way for a true IoT platform OS.
Add all these things up and I get the sensation that they lack the engineering base that is needed.
The bread and butter iPhone hasn't really seen any constant leaps for a flagship device sector. They are drip feeding features to users (many of which have been available on Android for years. This year there will finally be a design change it seems, but the Pixel comparisons are already out there and Apple knows that will only intensify if it turns out to be true.
iOS has implemented a lot of features from Android/HarmonyOS too over the last few years.
WWDC will be interesting. Will they do a Snow Leopard or make more big promises about upcoming features with no real shipping guidance?
What with catching up and trying to move several major projects forward, I can't help but think a few thousand more engineers are needed.
^Good analysis. It seems a lot of those protjects are also about extracting every penny out of the existing product lines, which is Tim Cook's supply chain specialty. But it's been a while since Apple has had a compelling new product launch. Vision could be one day, but it's not remotely as promising as other first gen efforts.
Tim Cook is either the most understanding boss, or Gianndrea has something on him. Siri has been terrible for years. Apple tries not to lay employees off, but someone failing at this level deserves to go. I'm sure he won't be hurting in retirement or at his next role.
Any one of us would be thrilled at that poorly perfuming chap' Golden Parachute.
I'm skeptical of a consumer-oriented robot being a desirable or successful product anytime soon.
However, building robots that can assemble iPhones and other products would be highly valuable to Apple. I think that should be their focus. I suspect that the lessons learned from building robots that can efficiently build iPhones would translate into other robot products for external customers.