Apple Finally Begins Focusing on Generative AI with New Job Openings and a Focus on Mobile

But, don't expect too much!

There are headlines everywhere about Apple CEO Tim Cook’s claims that the company has a giant R&D budget, and a “strategic” element of that investment has long been related to generative AI. Apple has a giant R&D budget, but how much of it has been applied to generative AI is likely less than the current spin would suggest. In and interview with Reuters, Cook commented:

"We've been doing research across a wide range of AI technologies, including generative AI, for years. We're going to continue investing and innovating and responsibly advancing our products with these technologies to help enrich people's lives. Obviously, we're investing a lot, and it is showing up in the R&D spending that you're looking at."

The Financial Times points out:

Apple is bulking up its expertise in generative AI to adapt it for iPhones and iPads, as the world’s biggest company by market value seeks to take advantage of the technology that has taken the industry by storm this year. 

The Cupertino-based tech giant is hiring for dozens of roles across offices in California, Seattle, Paris and Beijing that will work on large language models or LLMs — software that can produce plausible text, images or code in response to simple prompts.

All the jobs were advertised between April and July and indicated that Apple was working on “ambitious long-term research projects that will impact the future of Apple, and our products”.

Apple’s Generative AI Hiring Push

The Financial Times’ reporting is true. Between April and August (so far) in 2023, Apple has posted 18 jobs with Generative AI in the title or the first paragraph of the job description. Some of these include (emphasis mine):

• Proactive Intelligence, Applied Research Scientist — Generative AI (5 jobs)
  “AI represents a huge opportunity to elevate Apple’s products and experiences for billions of people globally. We are looking for Applied Research Scientists with a background and interest in Generative AI. You will be leveraging state-of-the-art Generative models to ship extraordinary products, services, and customer experiences for the iPhone, Mac, Apple Watch, iPad and more.”

• Senior Machine Learning Engineer - Large Language Models & Generative AI (1 opening)
  “The Intelligence Platform team empowers clients across Apple’s operating systems with high quality user-centric knowledge and inferences that enable next generation user experiences. We’re an applied Machine Learning team that leverages state of the art technologies like generative AI, graph machine learning, and private learning to deliver high quality inferences…By contributing to our team, you'll play an integral part in developing Siri, Photos, Music, and various other services, leaving a significant footprint on the evolution of our AI platforms.”

• Generative AI Applied Researcher - SIML, ISE (1 opening)
  “Are you excited about Generative AI? We are looking for experts in this space to join our applied ML R&D team at Apple! You will be inventing and shipping the next generation of these core technologies with a focused team. Our purpose is to surprise and delight users and developers worldwide. The team comprises domain experts in Computer Vision & Natural Language Processing (NLP) who contribute to a variety of shipping workflows you may already regularly use, including: Photos Search, Curation, Memories, Intelligent Auto-crop, Visual Captioning for Accessibility, Federated Learning on visual content, Real-time Classification & Saliency in Camera, Semantic Segmentation in Camera, and several on- device feature extractors across the system.”

• Sr Software Engineer - LLM Evaluation and Test Infrastructure (1 opening)
  “Do you believe generative models can transform smart assistants and creative workflows used by billions? Do you believe it can fundamentally shift how people interact with devices and communicate? We truly believe it can! We are looking for highly skilled senior software engineers experienced in architecting and deploying test and validation infrastructure for consumer-facing products using large-scale client/server ML systems. You will play a critical role in designing and building LLMs online and offline validation infrastructure, tools, frameworks, and methodologies that enable product teams across Apple to develop state-of-the-art generative AI solutions that power amazingly intelligent user experiences on iPhone, Mac, Watch, iPad, and more.”

• AI/ML Research Scientist (1 opening)
  “Are you passionate about building AI/ML products with focus on Augmented Reality (AR) and Virtual Reality (VR) applications? Are you passionate about Generative AI and large foundation models? Are you passionate about solving hard problems? Video Computer Vision org at Apple is looking for machine learning researcher to join the team of highly accomplished and deeply technical scientist.”

• ML Data Engineer, Technical Lead – SIML, ISE (1 opening)
  “Do you believe Machine Learning and AI can change the world? We truly believe it can! We are the Data Team of System Intelligence and Machine Learning. We are responsible for building high quality ML datasets at scale, used to train ML models that power AI-centric features for many Apple products (iPhone, iPad, Mac, Apple Watch and even AirPods). Such features go from the smart wallpaper on your iPhone Lock Screen, to the models that highlight the faces of your loved ones in your Photos app, to input experiences (eg autocorrect, next word prediction, handwriting recognition).”

• Several others…

A search for large language models brings up 23 jobs (some are redundant with the generative AI list). These were also posted since April 2023.

Siri-related jobs number 80, but many of those are NLU-based converational AI and do not mention generative AI. Some also were posted in the first quarter of 2023 and others are from 2022, though most are also since April.

Machine learning shows up in 360 jobs and if you broaden it to Apple’s category “Machine Learning and AI,” you find over 600 job listing results. Some of those have generative models or related generative AI nomenclature that don’t appear in the narrower “generative AI” search.

Apple is a large company. It always has technical roles listed for hiring. However, this does appear to be a concerted effort to bulk up generative AI staff capacity that has emerged since second quarter of 2023.

Devices First!

Apple sells devices first and foremost. The software that powers them is important, but it all comes back to the device. Software and applications exist in the service of delivering higher device sales. When viewed through that lens, it is unsurprising that Apple did not initially prioritize generative AI solutions. It was unclear how big of an impact it could have on device sales.

The long-standing Apple culture is also to work on its own timeline and not become distracted by the latest technology trends. Apple wants to set trends and lead culture and technology as opposed to follow.

The Apple Vision Pro is an attempt to do that. Virtual reality (VR) is an existing product category that has experienced slow adoption. That is the type of market Apple excels at disrupting. Generative AI can support those efforts. Apple will apply generative AI to both devices and services, but expect everything to be driven by what’s good for the device user experience. If it helps iPhone remain competitive and supports important features in the Vision Pro, it will get done.

Is Apple a Fast Follower?

I always receive negative comments when I say that Google is the most successful fast follower in history. “But wait,” I’m told. “There were music players before the iPod, smartphones before the iPhone, tablets before the iPad, and wireless earbuds before AirPods. Apple entered each market and established dominance.” This is supposedly a mic-drop comment on your social media network of choice. But, these commenters are missing the point?

My hypothesis, which is not entirely unique, is that Apple is not a fast follower at all. Its biggest hits have not been in established successful markets where it quickly followed the company that established market demand. Apple excels in markets that have yet to reach critical mass in adoption. It introduces a new product that finally delivers the value that drives demand. Think of it as first-to-product-market-fit at scale.

So, the hiring binge now is not about competing for existing generative AI segments. There are already three segments that have scale and happy users. Instead, I view the hiring signals as a recognition that Apple now expects to add generative AI features to existing products and services and would rather control the technology than rely on a third party.

What about Siri?

An obvious place where Apple could upgrade an existing solution with generative AI is Siri. It has hundreds of millions of regular users, but Siri is also not versatile. Its user base can be attributed more to a captive iPhone audience than exceptional product leadership.

Generative AI could make Siri more robust and compete more effectively with Alexa, Google Assistant, and the new kids on the block, such as ChatGPT and Bard. In addition, Alexa and Google Assistant are both expected to receive generative AI upgrades later this year and throughout 2024. Siri needs to go down the generative AI path to remain competitive.

However, Apple prioritizes control. Don’t expect it to let generative AI features run wild in Apple devices. Whereas software and services are subordinate to hardware in Apple-land, generative AI will be subordinate to both. Siri is a feature of the iPhone and generative AI will be a feature applied to both.

But There is AppleGPT

Bloomberg reported in July that Apple had already developed a large language model (LLM) but was unsure how to apply it to the company’s product ecosystem. You should not assume Apple’s LLM is on par with offerings from OpenAI, Google, and Meta. It certainly could, but the odds are “AppleGPT” is inferior. Other foundation models have been a focus of development and not a sideline.

Those LLMs and others have also had multiple iterations available to researchers, developers, or consumers. Making those models available has provided data and feedback to improve new model releases. Do you remember when Apple Maps debuted? Apple did not have enough data or feedback when it launched and it was a disaster. Iterations matter. GPT-3 came after GPT-2. GPT-4 after GPT-3.5, GPT-3, and ChatGPT. Google PaLM 2 came after PaLM 1 and LaMDA. Llama 2 is not Meta’s first rodeo.

Apple excels at working in private and surprising the world with new innovation. That approach is unlikely to work with generative AI.

On the Edge

If you want to forecast where Apple is likely to apply all of these machine learning scientists, you can look at devices, particularly edge computing devices. Apple’s industry contribution may be bringing generative AI capabilities into on-device computing environments. Google will be there first, and Meta is already working with Qualcomm. However, Apple has its own silicon chip development capabilities and could optimize edge-based generative AI computing the same way it has brought new performance capabilities to smartphones and laptops.

Apple’s reality distortion field around generative AI may finally be lifting, but don’t believe the hype that Apple has spent a lot on generative AI compared to its peers. AI and machine learning investments are related but not an exact match for generative AI.

Still, Apple has a lot of assets and could make a big impact. Just don’t expect that to be soon. Apple works on its own time table. And it this case it is not creating a market, it is playing for a niche.