Apple Intelligence is the intelligent approach to the AI race
An analysis of Apple's strategy for Apple Intelligence
Apple Intelligence is the intelligent approach to AI race
At WWDC 2024, Apple finally unveiled it’s long awaited AI efforts.
From the Wall Street Journal:
Apple joined the AI arms race, saying Monday it plans to bring a more personalized version of artificial intelligence to its 2.2 billion device users—including striking a deal with ChatGPT-maker OpenAI.
The new AI system, which it called “Apple Intelligence,” offered a preview to what many consider to be the holy grail of AI, a voice assistant empowered with enough personal user information to meaningfully help complete an array of tasks. Apple has partnered with OpenAI, and its ChatGPT, for some new AI functions, such as answering more complex queries or composing messages, capabilities that Apple’s AI can’t handle.
I highly recommend watching Apple’s 5 minute overview of Apple Intelligence, see below.
In this article, I’ll be sharing my analysis of the Apple Intelligence and covering the following topics:
Classic Apple branding
AI is a feature not a product
Siri’s second act
Apple’s privacy payoff
A tech trifecta — on-device, cloud, and ChatGPT
Classic Apple branding
Rebranding AI to Apple Intelligence is classic Apple marketing. Other clever rebrands include Ceramic Shield, Retina Display, and Spatial Computing. By rebranding a feature, Apple is able to shape a completely new narrative around an existing technology — one that is conveniently beneficial for Apple.
Apple is keenly aware that the public is fatigued by the constant barrage of Artificial Intelligence messaging. Not only does Apple Intelligence cleanly cuts through this noise, it also leverages Apple’s well established brand halo and positions itself as the personal and accessible approach to AI.
This counter positioning is brilliantly encapsulated in Apple’s recent advertising, see image below, with the tagline “AI for the rest of us”.
Fun fact: The tagline is actually a throwback to the original tagline from the 1984 launch of the Macintosh.
AI is a feature not a product
While fellow tech giants Google and Microsoft have doubled down on chat bots, and startups like Humane and Rabbit have unsuccessfully poured millions into AI devices, Apple, on the other hand, clearly views AI as a feature rather than a product.
Apple’s strategy for Apple Intelligence focuses on augmenting and streamlining practical tasks that billions of its users encounter daily.
Some of the features include:
Better photo editing, such as object removal.
AI generation of custom emojis using text prompts called Genmoji.
Writing Tools that can automatically help users can rewrite, proofread, and summarise text.
An enhanced Mail app that can prioritise and summarise emails, and suggest replies.
Image Playground lets users create generative images , with three styles: Animation, Illustration, or Sketch, to pick from.
Enhanced Siri capabilities, which I’ll go into more detail in the section below.
ChatGPT access integration across Siri and Writing Tools.
Siri’s second act
For most users, Siri hasn’t lived up to its promise of being a true virtual assistant. However, with Apple Intelligence, Siri might finally fulfil its true vision from 12 years ago.
During the keynote, it was announced that Siri’s capabilities will be upgraded with the following:
Personal context awareness: Because Siri has direct access to the personal information on your Apple device, you can now ask questions like “When does dad’s flight land?” or “What was the apple pie recipe that John shared with me?”.
Better natural language understanding: Siri can understand you even if you make mistakes, see example below. It can also understand information from a previous request.
Product knowledge: Siri can now act as your personal tech support for Apple devices. In the screenshot below, Siri provides step by step instructions on how you can schedule a text message to send later.
If — big emphasis on if — Siri can live up to its new capabilities and shrug off its current perception of clunkiness, we might just see Siri become the default way that users interact with their Apple devices — à la Samantha from the movie Her.
Apple’s privacy pay off
Apple has continually demonstrated its strong commitment to user privacy through initiatives such as the roll out of App Tracking Transparency (ATT) and the company's famous refusal to comply with the FBI's request for a backdoor to the iPhone.
This investment in privacy will pay off in spades when it comes to users adopting Apple Intelligence, especially since a key concern of consumers is the misuse of data in AI. According to a Pew Research report, 70% of Americans say they have little to no trust in companies to make responsible decisions about how they use AI in their products.
It certainly helps that Apple’s crack marketing team can craft slick messaging that proselytises Apple Intelligence’s privacy features — see the video below.
A tech trifecta - on device, cloud, and ChatGPT
Apple has taken a three-pronged approach to its AI models, combining on-device processing, cloud computing, and ChatGPT integration.
With Apple’s staunch commitment to privacy, it’s no surprise that Apple’s ~3 billion parameter language model runs locally on the device. What is surprising to me is Apple’s claim that their model outperforms other open-source models such as Microsoft’s Phi-3-mini and Google’s Gemma-7B. See images below.
The key challenge with relying on an on-device model is its inability to answer more complex requests. Apple’s solution to this is the creation of Private Cloud Compute.
From Apple:
To run more complex requests that require more processing power, Private Cloud Compute extends the privacy and security of Apple devices into the cloud to unlock even more intelligence.
With Private Cloud Compute, Apple Intelligence can flex and scale its computational capacity and draw on larger, server-based models for more complex requests. These models run on servers powered by Apple silicon, providing a foundation that allows Apple to ensure that data is never retained or exposed.
Even with Private Cloud Compute, users are likely to send requests that require significant processing power or specialised knowledge. This is where ChatGPT integration comes in. Apple Intelligence will delegate the most difficult tasks to ChatGPT but only after the user actively grants permission. This partnership also conveniently shifts any privacy issues and blame to ChatGPT.
Apple’s combination of on-device and cloud models isn’t new; in fact, both Google and Microsoft have similar strategies in place. What sets Apple apart is their emphasis and commitment to privacy. This user trust will likely result in higher adoption rates of Apple Intelligence compared to other technology companies. After all, Google has built its fortune by selling user data to advertisers.
Why Apple Intelligence could fail
Generally speaking, I’m cautiously optimistic about Apple Intelligence’s chances of success. However, I would be a crap analyst if I didn’t point out a few risks that could cause it to fail.
Apple fails to execute what it showcased at WWDC 2024.
The emergence of a large language model (LLM) that is overwhelmingly powerful, practical, and personal — leaving consumers no option but to switch.
Another device maker, i.e. Google Pixel, integrates AI more effectively and challenges iPhone's market dominance.
Closing thoughts
I firmly believe the technology company that will win the AI race will be the one that meets two key criteria:
Deploys an AI offering with the least friction for users
Achieves the widest distribution
Based on the announcements at this year’s WWDC, Apple appears poised to meet both requirements and might just take the lead in the AI race.
Things that caught my attention
FAA investigates counterfeit titanium used in some Boeing and Airbus jets — yikes.
Coffee YouTuber and World Barista Champion, James Hoffmann, Drank A $335 (£265) Espresso. Watch his video to see his review of it.
- from the has released an in-depth Guide to positioning.