Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

MacRumors

macrumors bot
Original poster
Apr 12, 2001
68,543
39,399


One of the major factors that contributed to Apple's heavy focus on artificial intelligence in iOS 18 was an experience that software chief Craig Federighi had with GitHub Copilot, according to The Wall Street Journal.

craig-federighi.jpg

In a report on Apple's AI ambitions, WSJ says that Federighi became an AI "convert" after testing the Copilot coding tool introduced by Microsoft-owned GitHub. Copilot, which uses OpenAI technology, was created to help programmers write, troubleshoot, and translate code into different programming languages.

After "playing around" with GitHub Copilot in December of 2022, Federighi explained that he had "come to appreciate" generative AI, and he reportedly tasked engineering employees with creating new ways to integrate AI features into Apple's apps and features. Prior rumors have suggested that Apple is aiming to integrate AI into as many apps as possible with iOS 18 and future software updates.

Apple accelerated development on internal generative AI development, even recruiting some former Apple Car employees to work on the technology, but it has also held discussions with OpenAI and Google and is expected to rely on both its internal AI models and external partnerships.

Rumors about features coming in iOS 18 indicate that Federighi has been successful in incorporating AI into the operating system in a number of ways, with Apple planning to add AI features to Messages, Mail, Photos, and other key apps. Siri will also be overhauled with generative AI, making the personal assistant smarter and more capable.

A rundown of all of the new features rumored for iOS 18 can be found in our roundup.

Article Link: Apple's Push Into AI Allegedly Happened After Craig Federighi Tried Microsoft AI Coding Tool
 
Last edited:
This claim seems a little suspect, at least in terms of the timing. Are they suggesting that Apple designed and built M4 with a heavy AI emphasis, based on stuff that happened in 2022? For a chip shipping in volume in actual products in early 2024 (meaning production of the chip starting in 2023), one would have expected them to have started before 2022.
 
Lmao so they said: we need to start to do (copy) something after (one of the leadership) used a public released consumer facing product?

Did no one saw the potential future of AI further back when generative AI was still in labs? And only people as high as Craig able to steer the ship?

How about try some Chinese EVs, especially their eco system integration and make decision to go all in on Apple Car. Or try smart home solutions from Mijia and re-org their Apple Homes?

That’s true innovation

Are Apple employees even ready to go all in on AI? It needs years of researches, the PMs need to think in a completely different way, and softwares needs major rearchitect.

I guess we’ll see in a few days, but this is concerning..
 
Last edited:
This claim seems a little suspect, at least in terms of the timing. Are they suggesting that Apple designed and built M4 with a heavy AI emphasis, based on stuff that happened in 2022? For a chip shipping in volume in actual products in early 2024 (meaning production of the chip starting in 2023), one would have expected them to have started before 2022.

There's about 18 months between December 2022 and May 2024. M4 was probably well into the design cycle but it shouldn't have been too late to make changes to focus on NPU.

If Apple had stepped on the NPU gas earlier, Apple should be occupying the bright orange bar.

AMD COMPUTEX CLIENT PRESS DECK-01-01 (31).png
 
There's about 18 months between December 2022 and May 2024. M4 was probably well into the design cycle but it shouldn't have been too late to make changes to focus on NPU.

If Apple had stepped on the NPU gas earlier, Apple should be occupying the bright orange bar.

View attachment 2385825
The design likely would have had to be finished by mid 2023. I'm not saying this is absolutely impossible, but it just seems unlikely and uncharacteristic for Apple.

Maybe what the article says is true for the software and OS side, but there is a good chance the hardware design was already in the works before that.
 
Where would Apple be without Google, Samsung and Microsoft? Not only is competition good, but for Apple’s future, it’s essential it seems. There was a time long ago when Apple seemed to lead in the talent department. Now, Craig needs to see what the other guy is doing and hope they can copy and make enough changes so the lawsuits don’t come rolling in.
 
This is a PR puff piece. No company of any decent size, and especially Apple's size, works like this. A single executive can't do this stuff. It requires a lot of approval, resources, unity among other execs and lower level staff and more.

I can believe he loved using GH Copilot, and wanted to explore it more. But anything past that point hits into stage gating concepts and ideas into reality, and at Apple's size, that takes a lot of time.
 
This is a PR puff piece. No company of any decent size, and especially Apple's size, works like this. A single executive can't do this stuff. It requires a lot of approval, resources, unity among other execs and lower level staff and more.

I can believe he loved using GH Copilot, and wanted to explore it more. But anything past that point hits into stage gating concepts and ideas into reality, and at Apple's size, that takes a lot of time.

A senior VP who has worked at Apple for the past 30 years and reports directly to Cook, can't do this? Tim literally calls Craig, "Superman." This is Silicon Valley, not ExxonMobil.
 
This claim seems a little suspect, at least in terms of the timing. Are they suggesting that Apple designed and built M4 with a heavy AI emphasis, based on stuff that happened in 2022? For a chip shipping in volume in actual products in early 2024 (meaning production of the chip starting in 2023), one would have expected them to have started before 2022.
Apple’s been upping the ante on its Neural Engine NPUs for years, so the improvements to the M4 were likely planned no matter what. The biggest jump actually came with the M2.

And keep in mind the M4 isn’t as big of a leap as it might seem based on advertising. TOPS ratings are largely useless unless specifying what specific operation you’re counting. The M3’s 18 TOPS is based on FP16, whereas the M4’s 38 TOPS using INT8. The two aren’t entirely comparable but FP16 can essentially do more work per operation at greater precision. Based on preliminary testing the M4’s NPU is about 5% faster in real world operations than the M3’s.
 
The more inside baseball we get, the more I realize lead times aren't what I expected. That Craig only started playing with Copilot for Visual Studio last year and then was like we should do this is not at all how I expected their machine learning development to have gone, I thought they would have been working on something years back and were just later to deploying it.
 
The more inside baseball we get, the more I realize lead times aren't what I expected. That Craig only started playing with Copilot for Visual Studio last year and then was like we should do this is not at all how I expected their machine learning development to have gone, I thought they would have been working on something years back and were just later to deploying it.

As others have already said: I don’t buy this at all. There’s no way it played out like the article suggests.
 
Kind of pathetic. It's like great grandpa being introduced to a microwave while it has already become people's daily tool for a long time.
Not sure how this is pathetic. Copilot just officially launched in mid 2022 so it wasn't like he tried it years after people had been using it. Not to mention a lot of the other AI features in the past were often technically impressive but not all that useful, so there was a good reason to be skeptical. LLMs have enabled some of the first use cases that are actually somewhat useful for everyone.
 
There's about 18 months between December 2022 and May 2024. M4 was probably well into the design cycle but it shouldn't have been too late to make changes to focus on NPU.

If Apple had stepped on the NPU gas earlier, Apple should be occupying the bright orange bar.

View attachment 2385825

Sincere question - who produced that graphic chart? Seems to be taken from a powerpoint or similar prezzo.

Edit later - clearly it came from AMD, with the trademark "TM" superscript alongside their name & product :rolleyes:
 
Last edited:
"...successful in incorporating AI into the operating system in a number of ways... Siri will also be overhauled with generative AI, making the personal assistant smarter and more capable..."

The bar there was pretty low, though. I'm keeping the hope alive!
 
  • Like
Reactions: rehkram
One of the major factors that contributed to Apple's heavy focus on artificial intelligence in iOS 18 was an experience that software chief Craig Federighi had with GitHub Copilot, according to The Wall Street Journal.
I think the press likes to create stories like this to make fun of executives.
Apple's AI chief John Giannandrea is leading development on large language models within Apple, and Giannandrea reports directly to Apple CEO Tim Cook. Giannandrea established a team that works on conversational AI four years ago, and work has since ramped up.
That would be 2020 going by that not Apple paying attention to the below observation in 2022.
With the 2022 debut of OpenAI's ChatGPT, chatbots suddenly became the must-have feature. Microsoft and Google have both launched chatbots, but there are so far no signs that Apple has a consumer-oriented product launching in the near future.
 
Last edited:
If he just started with Copilot last year, then that means it's going to be another 4 years before a hardware product with a new chip is available that was designed around it.

Chips generally take 5 years from concept to consumer.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.