Revamping Siri: Apple's A.I. Revolution

14 min read        

So let it be said that Apple is playing catch up with AI. That said, the company is moving quickly to revamp Siri with powerful generative A.I, so that they are running in parallel with their chatbot competitors, or better yet, surpass them. With the biggest company shake-up in a decade, Apple is set to revolutionize how we interact with our iPhones. What will this bold new direction mean for Apple and A.I.? This will be the company's most significant reorganization in a decade.

Welcome to That's Life, I Swear. This podcast is about life's happenings in this world that conjure up such words as intriguing, frightening, life-changing, inspiring, and more. I'm Rick Barron your host.

That said, here's the rest of this story

Time to play catch up

Apple is expected to show off its AI work at its annual developer's conference on June 10, 2024. It's expected to release an improved Siri. This is the company's most intense and widespread endeavors to respond and pivot to the AI frenzy sweeping the technology industry.

So how did all this come about?

According to two individuals familiar with Apple's operations, the tech giant's top software executives, Craig Federighi and John Giannandrea, reached a pivotal decision early last year: Siri, the company's virtual assistant, required a complete overhaul.

This decision came after the executives spent weeks extensively testing ChatGPT, the groundbreaking chatbot developed by OpenAI. The product's impressive utilization of generative artificial intelligence, capable of composing poetry, generating computer code, and answering intricate queries, rendered Siri obsolete.

Introduced in 2011 as the original virtual assistant integrated into every iPhone, Siri had been limited to handling individual requests for years and could not sustain coherent conversations. It frequently misunderstood questions posed by users. In contrast, ChatGPT demonstrated a remarkable capacity to comprehend context, and with little effort, grasping that if a user inquired about the weather in San Francisco and subsequently mentioned, "What about New York?", it knew the user sought another weather forecast for the latter location.

Apple has some catching up to do. The company largely sat on the sidelines when OpenAI's ChatGPT took off like a rocket last year. It watched as Google and Microsoft Corp. rolled out generative AI versions of their search engines, which spit out convincingly human-like responses to users' queries. Microsoft also updated its Windows apps with smarter assistants, and Amazon unveiled an AI-enhanced overhaul of Alexa.

Chief Executive Officer Tim Cook says  Apple has worked on generative AI technology for years. It's plain to see that in no uncertain terms, Apple executives were caught off guard by the industry's sudden AI fever and have been scrambling since late last year to make up for lost time.

The realization that new technology had surpassed Siri's capabilities set in motion Apple's most significant reorganization in over a decade. Determined to pick up the pace and enter the tech industry's race for artificial intelligence dominance, the company has elevated generative AI to a "tent pole project" – an internal designation reserved for once-in-a-decade initiatives around which employees are organized and mobilized.

I like the "tent pole project" designation. Why?

It reminds me of a time where I had the opportunity working at Apple, and such a ‘hit the decks running’ moment happen. That moment was where I was one of several supervisor’s working at a factory in Fremont, California. It was inside that building we were manufacturing the very first Macintosh computer. I remember Steve Jobs telling the team at the factory in Fremont, that this was a make-or-break moment for Apple.

Apple is anticipated to unveil its advancements in artificial intelligence at its annual developers' conference on June 10. During this event, the company is expected to release an improved version of Siri, its virtual assistant, which will boast enhanced conversational abilities and versatility.

The update to Siri is at the forefront of Apple's broader endeavor to embrace generative AI across its business operations. As part of this initiative, the company is increasing the memory capacity of this year's iPhones to support Siri's new capabilities. Furthermore, Apple has explored licensing complementary AI models that power chatbots from several prominent companies, including Google, Cohere, and OpenAI.

Apple's leadership team harbor concerns that the emerging advancements in artificial intelligence technology pose a threat to the company's dominance in the global smartphone market. Their apprehension stems from the potential of this new technology to evolve into a primary operating system, effectively displacing the iPhone's iOS software.

Moreover, this groundbreaking AI technology can create an ecosystem of AI-powered apps, referred to as "agents," capable of performing tasks such as ordering rides or scheduling appointments. This prospect directly undermines Apple's App Store, generating a staggering $24 billion annual sales.

Apple's fears extend beyond the realm of software, as the company is wary that a failure to develop its own AI system could render the iPhone a mere "dumb brick" compared to other technological advancements. Despite the uncertainty surrounding the number of regular Siri users, the iPhone currently captures an impressive 85 percent of global smartphone profits and generates over $200 billion in sales.

This profound sense of urgency has contributed to Apple's decision to cancel its ambitious $10 billion project to develop a self-driving car, reassigning hundreds of engineers to focus on advancing artificial intelligence initiatives.

Executive Sponsors

What is the new strategy to spearhead Apple's drive in A.I.?

Apple had disclosed its development of a proprietary large language model, codenamed "Ajax," and the rollout of an internal chatbot, dubbed "Apple GPT," to test its capabilities back in July of 2023, as initially reported. The critical juncture now lies in determining whether the technology meets the competition and how Apple plans to integrate it into its product offerings.

John Giannandrea and Craig Federighi, the company's senior vice presidents overseeing AI and software engineering, are spearheading this endeavor. Within Tim Cook's leadership team, they are referred to as the "executive sponsors" of the generative AI push. According to sources, Eddy Cue, the head of services, is also involved in this initiative. The trio is currently on track to allocate approximately $1 billion annually towards this undertaking.

John Giannandrea

John Giannandrea. Courtesy of: CNBC

Giannandrea is overseeing the development of the underlying technology for a new AI system, and his team is revamping Siri in a manner that will deeply integrate it. This smarter version of Siri could be ready as early as next year, but there are still concerns surrounding the technology, and it may take longer for Apple's AI features to permeate across its product line.

Craig Federighi

Craig Federighi. Courtesy of: Forbes

The software engineering group under Federighi's leadership is incorporating AI into the next iteration of iOS. There is a directive to infuse it with features powered by the company's large language model (LLM), which leverages vast data to refine AI capabilities. These new features should enhance the ability of both Siri and the Messages app to field queries and auto-complete sentences, mirroring recent advancements in competing services.

Concurrently, Apple's software engineering teams are exploring the integration of generative AI into development tools like Xcode, a move that could potentially accelerate the process of writing new applications for app developers. This would align with services such as Microsoft's GitHub Copilot, which offers auto-complete suggestions to developers as they write code.

Eddy Cue

Eddy Cue. Courtesy of: 9to5 Mac

Meanwhile, Cue's organization is actively pushing to incorporate AI into as many apps as possible. The group is investigating new features for Apple Music, including auto-generated playlists (a capability that Spotify introduced earlier this year in collaboration with OpenAI) and the company's productivity apps.

Cue's team is also examining how generative AI can be utilized to assist users in writing tasks within apps like Pages or auto-creating slide decks in Keynote. Once again, this mirrors Microsoft's features that have already been launched for its Word and PowerPoint apps. Additionally, Apple is testing the implementation of generative AI for internal customer service applications within its AppleCare group.

Apple has explored the possibility of creating servers powered by its iPhone and Mac processors. This strategic move could yield cost savings for the tech giant while fostering consistency between the tools utilized for processes in the cloud and those on its devices.

Rather than directly competing with ChatGPT by releasing a chatbot capable of tasks such as poetry writing, three sources familiar with Apple's efforts stated that the company has focused on enhancing Siri's performance in handling tasks it already undertakes. These include setting timers, creating calendar appointments, adding items to grocery lists, and summarizing text messages.

Apple plans to market the improved Siri as a more private alternative to rival AI services, as it will process requests directly on iPhones rather than remotely in data centers. This strategy is expected to yield cost savings as well. In contrast, OpenAI incurs approximately 12 cents for every 1,000 words generated by ChatGPT due to the costs associated with cloud computing.

Apple faces potential risks while pursuing a strategy that relies on a smaller AI system housed within iPhones rather than a larger one stored in data centers. Research has indicated that smaller AI systems could be more prone to making errors, known as "hallucinations," than their larger counterparts.

Tom Gruber, a co-founder of Siri who worked at Apple until 2018, provided insight into this challenge, stating, and I quote: "It's always been the Siri vision to have a conversational interface that understands language and context, but it's a hard problem. Now that the technology has changed, it should be possible to do a much better job of that. So long as it's not a one-size-fits-all effort to answer anything, then they should be able to avoid trouble." End quote

Despite the hurdles, Apple possesses several advantages in the AI race. With over two billion devices globally, the company has a vast platform for distributing AI products. Additionally, Apple boasts a leading semiconductor team producing sophisticated chips capable of powering AI tasks such as facial recognition.

Over the past decade, Apple has faced challenges formulating a comprehensive A.I. strategy, with Siri seeing minimal improvements since its launch. This shortfall has diminished the attractiveness of Apple's HomePod smart speaker, as Siri often struggles to perform basic tasks like responding to song requests effectively.

According to John Burkey, a former Siri team member who now heads the generative A.I. platform Brighten.ai, the Siri team has not received the same attention and resources as other Apple divisions. Apple's various departments, including software and hardware, operate independently with limited information sharing. However, it must be seamlessly integrated across products for A.I. to thrive.

An ongoing deliberation within the company revolves around the deployment strategy for generative AI: whether to implement it as an entirely on-device experience, a cloud-based setup, or a hybrid approach combining both. An on-device approach would offer faster performance and enhanced privacy safeguards, but leveraging Apple's LLMs via the cloud could enable more advanced operations.

However, the on-device strategy presents challenges in updating the technology and keeping pace with the rapidly evolving industry. Considering this, it would not be surprising if the tech giant adopted a combined approach, utilizing on-device processing for certain features while relying on cloud computing for more advanced tasks.

Recruitment and Secrecy

Apple has faced difficulties in attracting and retaining top A.I. researchers. While the company has acquired A.I. firms led by industry experts, these leaders have typically departed within a few years.

Various factors contribute to these departures, with Apple's culture of secrecy being significant. Unlike Google, Meta, and Microsoft, Apple publishes fewer A.I. research papers and is less active at conferences.

Ruslan Salakhutdinov, a prominent A.I. researcher who left Apple in 2020 to return to Carnegie Mellon University, explained, "Research scientists consider their options: Can I return to academia or join a research institute where I can work more openly?"

In recent months, Apple has increased its publication of A.I. papers. However, leading A.I. researchers have expressed skepticism about their value, suggesting these publications aim to create an impression of substantial work rather than showcase potential market innovations.

Tsu-Jui Fu, an Apple intern and A.I. doctoral student at the University of California, Santa Barbara, authored one of these recent papers. Last summer, he developed a system that edits photos using written commands instead of traditional Photoshop tools. Although Apple provided the necessary GPUs for his project, he had no direct interaction with the A.I. team working on Apple products.

Despite interviewing for full-time positions at Adobe and Nvidia, Fu plans to return to Apple after graduation, believing he can make a more significant impact there. He noted, "A.I. product and research is emerging at Apple, but most companies are very mature. At Apple, I have more room to lead a project instead of just being a team member."

What can we learn from this story? What's the takeaway?

Apple was caught flat-footed when ChatGPT and other AI tools took the technology industry by storm. However, the company is now preparing its response and plans to develop features for its full range of devices.

When it comes to getting this right, the stakes are high. Generative AI has quickly become much more than a buzzword and will be central to the next several decades of computing. Apple knows it can't afford to take a back seat

Well, there you go, my friends; that's life, I swear

For further information regarding the material covered in this episode, I invite you to visit my website, which you can find on Apple Podcasts, for show notes calling out key pieces of content mentioned and the episode transcript.

As always, I thank you for the privilege of you listening and your interest.

Be sure to subscribe here or wherever you get your podcast so you don't miss an episode. See you soon.