What's Apple's AI Play?
Edge device King Apple is being quiet in the AI Race ... too quiet?
TL;DR - Whether Apple makes, buys or partners to make it happen, the look and feel of their products will need to change significantly to stay current with the AI platform shift.
All Quiet on the Apple AI Front
While AI goes through its “iPhone moment”, as NVidia’s CEO put it, the maker of the iPhone has been rather quiet about AI lately. In contrast to the very aggressive approach Microsoft is taking, and the AI-everything Google approach, Apple’s executive commentary around AI has been cautious.
From Statistica, this chart shows the AI mentions in the latest earnings calls for Big Tech Companies. Clearly, Google/Alphabet, Meta and Microsoft are much more vocal in highlighting their AI investments and developments than Amazon and Apple.
At least one commentator seems to think Apple’s caution may leave them behind, saying Apple's AI strategies, especially for Siri, aren't very smart:
Siri was introduced to the world in 2011 – 12 years ago – and was lauded at the time as a breakthrough in AI technology. Because it was. However, over the years, Siri has failed to live up to its potential. While the technology has improved since its inception, the AI efforts of Google and OpenAI have greatly eclipsed it.
Apple’s Siri, similar to Amazon’s Alexa, is a one-time cutting-edge leader with ‘canned query and response’ chat capabilities were ground-breaking for the 2010s, but that now feel dated in the era of chatGPT. It’s not close to what advanced LLMs can do.
Some iPhone users have been able to hack getting GPT-4 connected into Siri, but it begs a question of why Apple won’t lead and integrate it or something like it themselves.
It also raises the question of Who will win Big Tech’s race for the best AI assistant: Google, Apple, Meta or Amazon? From the Wall Street analyst perspective, Google seems to have the all-around technology advantage, while Apple is “positioned to deliver a broad-based personal AI assistant, but AI model capabilities are less clear.”
Hardware is Apple’s advantage. With about 2 billion devices — roughly 1.2 billion of those iPhones — and most compatible with Siri, Apple delivers more edge devices to consumers than anyone and has a ton of user data.
The big question mark is lack of performance data on Apple’s AI capabilities. “Little has been said around whether Apple is building its own LLMs, or if they’ll decide to partner with an open-source vendor such as StabilityAI,” Morgan Stanley said.
Apple’s AI Pedigree
"We see enormous potential in this [AI] space to affect virtually everything we do... It will affect every product and every service that we have." - Tim Cook, Apple CEO, February 2023
All of Apple’s current products have some AI capabilities and could be described as AI tools. The iPhone, which accounts for the majority of Apple’s revenue, is the company’s most prominent AI device.
Apple first incorporated AI into the iPhone in 2011 when it introduced the speech-based personal assistant Siri. As we mentioned in a prior article, Siri uses deep learning AI models to recognize and process speech and accurately deliver results in a clear and recognizable synthesized voices.
Apple has been putting their own processors onto their devices for several generations, and they have incorporated GPUs and AI-specific processors called Neural Engines into them. Apple’s Neural Engines are optimized for highly parallel computing to perform AI model inference. This can address voice recognition, text translation, face and object recognition, OCR (optical character recognition), and a number of other natural language processing, computer vision and AI tasks.
Their latest generation M2 processor, announced in June 2022, combines CPU (8-CPU core), GPU, and Neural Engine. The M2 Neural Engine can process up to 15.8 trillion operations per second (TOPS). These chips make the Mac among the most highly capable computers for running generative-AI applications.
Apple’s Neural Engine is also found in the chips powering Apple’s iPhone. iPhone users might not be aware of the technology under the hood, but experience the smoother performance of AI features nonetheless.
Many of these AI features are small AI models doing minor tasks to improve the user experience. They are quiet but pervasive.
That is Apple’s strength: Creating great user experience by making the technology user-friendly to the point where it is invisible.
Cautious with the LLMs
Invisible is one thing, but AWOL is another. Is Apple’s cautious and ‘thoughtful’ approach sufficient to dealing with rapidly improving LLMs and the AI tsunami?
The user experience is the reason Apple will have to come to terms with LLMs quickly. As I put it in my “Fundamental Thoughts on AI” article, the prompt is the interface now.
Apple is the King of edge devices right now. The release of chatGPT and plug-ins for it enables a better way to interact with software and upends all prior interfaces. Natural language, whether through chat or voice, should be the interface for all tools and applications.
All edge devices need to adapt to this new reality. They need to fully support an AI-enabled natural language interface for all their key applications. The OS itself needs to become natural-language capable, perhaps going as far as embedding an LLM in the OS itself.
Apple needs to take note. Apple understands that this generative AI still has trust and safety issues, so they are unlikely to release LLM-based interface technology unless it believes it’s trustworthy enough to bring to its billion-plus customers.
While Apple is out of the LLM race publicly right now, hopefully they are working on how to build this into their products. They have many options to get in. Apple likely won’t make a generic large language model AI system like ChatGPT or Google Bard, nor need to. It could partner with others, customizing an LLM based on its intimate knowledge of its customers.
Like it has bought into AI technology before, Apple could even acquire to gain capability. Over the years, Apple has acquired a number of companies to push boundaries in machine learning and AI, and from 2016 to 2020, Apple was the leading buyer of AI companies.
Whether Apple makes, buys or partners, the look and feel of their products will need to change to stay current with the AI platform shift.
Apple’s Recent AI Announcements
While the rise of LLMs is a challenge, Apple is far from an also-ran in the AI race. Apple has been using machine learning and AI for years, they have AI enablement in their silicon, key AI applications on their iPhones, and are continuing to develop AI applications, furthering the use of AI for better user experiences.
Some recently announced AI developments by Apple include:
Apple is developing an AI-powered health coaching service code named Quartz. “The tech giant is reportedly also working on technology for tracking emotions and plans to roll out an iPad version of the iPhone Health app this year.”
Apple is developing a Personal Voice capability. “Using Personal Voice, an individual will be able to use an Apple Silicon-equipped Mac, iPhone or iPad to create a voice that resembles their own. If the ability to speak is lost, text the user generates on the device can then be converted to voice, for use in a variety of ways.”
Another is Apple’s work on self-driving cars. Due to the companies infamous secrecy, there is little to reveal about where they are at, except to say that it has been a big AI challenge for many companies.
Finally, there is the upcoming launch of their AR headset in June. AR stands for “Augmented Reality” which means it overlays virtual elements on your real-world view. This headset allows both VR (virtual reality) and AR (augmented reality) modes.
Apple’s AR headset would be a whole new product category, the first since Apple launched the Apple Watch almost 10 years ago. (Just so you don’t get too excited, the initial pricing on these are going to be around $3,000.)
Why is this important in the context of AI? While not an AI development per, this AR headset platform has the potential to synergize with AI-driven natural language interfaces to create a vastly more capable user experience.
Apple is bringing out a whole new device category, with a whole new platform and experience. For this to succeed, it will need the best user interface, which is an AI-driven voice-enabled natural language interface.
As a leader in edge devices, including smart phones, laptops, PCs, and soon AR headsets, Apple has differentiated by enabling AI on the edge. Yet AI is not standing still. While Apple will be “thoughtful” and less vocal than others, Apple will need to move deliberately and swiftly in incorporating the latest AI models and capabilities to keep their products on the cutting edge.
And if they do that, the consumer wins.