Apple’s AI Evolution: Embracing Generative AI with a Privacy-Centric, User-Friendly Approach

To take another look at this article, go to My Profile and then to Saved Stories to view it again

Purchasing through the links in our articles could result in us receiving a commission. Find out more.

Steven Levy

Regardless of the presentation, Apple's prowess in Artificial Intelligence (AI) stands undeniable. While tech behemoths like Google, Microsoft, and Amazon have been vocal about their AI endeavors, Apple maintained a veil of silence until its leaders decided to open up. During an exclusive preview, I learned that Apple aimed to dispel the notion of being behind in the race of this crucial technological era. Craig Federighi, the head of software, Eddie Cue, the overseer of services, along with key innovators, made a case for Apple's longstanding leadership in AI, albeit without the usual fanfare. They highlighted that sophisticated machine learning techniques have been integral to several of their products for years, promising further enhancements, especially in Siri. Uniquely, Apple's commitment to data security meant that its AI projects were marked by stringent privacy safeguards. When questioned about the size of Apple's AI team, Federighi's response was simply, "A lot." Another leader pointed out Apple's cautious approach towards AI, deliberately steering clear of the more speculative and sensational aspects pursued by others in the field. They stressed that Apple's strategy in AI development would reflect the company's distinctive ethos.

Eight years back, the buzz in the tech world centered around deep learning AI technology. However, just a year later, an innovation known as Transformers emerged, leading to the creation of a new type of intelligent software known as generative AI. This technology underpinned the revolutionary ChatGPT developed by OpenAI. Suddenly, the benchmark for tech companies became their eagerness to embrace this new trend. Competitors of OpenAI quickly took notice and adapted. Apple, however, lagged behind. A significant number of its top AI researchers had been dedicating their efforts to developing autonomous vehicles and the costly Vision Pro mixed-reality headset. But over the past year, Apple shifted its focus, redirecting its experts from those endeavors to develop a strategy for generative AI. This pivot was showcased at the Worldwide Developers Conference this week, where Apple unveiled its new direction.

In a departure from the norm, the spotlight during the recent Apple event wasn't on product launches but rather on Apple's commitment to generative artificial intelligence (AI). CEO Tim Cook shed light on this unusual focus in a post-keynote interview, stating, "The audience was particularly curious about our stance on generative AI." However, reminiscent of their cautious approach back in 2016, Apple is set to explore generative AI, but in a manner that's distinctively theirs. The tech giant has opted not to categorize its innovation under the generic term of artificial intelligence. Instead, it has introduced the term "Apple Intelligence," a novel designation aimed at distancing the brand from the more daunting implications of this advanced technological trend. Apple's vision does not align with chasing after the concept of the singularity or bringing to life scenarios depicted in the film "Her." The company intends to leverage this cutting-edge technology to bolster productivity and creativity. By infusing AI with Apple's unique touch, the company aims to make this formidable technology more approachable and user-friendly.

The timing for this strategy is impeccable. I consider the era of generative AI to have commenced with the launch of ChatGPT in November 2022. Throughout 2023, we've been grappling with its implications, and many are now experiencing a backlash. There's a widespread discomfort with the so-called "hallucinations" produced by AI, alongside fears of job displacement. Moreover, the majority are still in the dark about how AI could benefit them. Moving into 2024, forward-thinking companies are zeroing in on practical applications for this astonishing technology. Apple has adopted the mantra "AI for the rest of us," a phrase that made a singular appearance during their keynote. This choice of words deliberately echoes the marketing of the original Macintosh, suggesting Apple's intention to democratize AI in a manner akin to how it popularized the graphical user interface with the Mac.

In sharp contrast to their lofty aspirations, the innovations Apple showcased in their presentation fell short of being groundbreaking. Many of the demonstrations revolved around basic tasks such as summarizing texts, transcribing speech, automating email responses, managing email clutter, generating text from prompts, and removing unwanted figures from photos. These functionalities are now expected in the age of generative AI. Apple's argument is that it seamlessly integrates these enhancements into everyday tasks, making them not only useful but also enjoyable for users. Additionally, Apple has introduced some clever features in these offerings. For instance, the Photos app now has an enhanced search function powered by AI, capable of understanding the content and subjects of photos to facilitate searches based on broad queries. When it comes to crafting automatic responses to emails, Apple's system may prompt the user with a straightforward question—such as confirming a meeting and suggesting times—with a single click, thereby generating an appropriate reply that aligns with the user's preferences. Importantly, given the vast amount of personal information stored on Apple devices, the company's AI leverages this data to produce pertinent results while ensuring it remains encrypted on the device to safeguard privacy. Apple SVP Federighi emphasized this approach as offering “intelligence that understands you.” Furthermore, Apple has pledged to engage external auditors to ensure the data's security is as promised.

Apple's latest announcements have stirred considerable interest, particularly around the updates to its digital assistant, Siri, which has seemed outdated compared to the advancements in generative AI. Apple has made a bold promise that by potentially 2025, Siri will not only improve in engaging in conversations but will also evolve into an exceptionally capable assistant capable of handling intricate tasks across various apps. This ambitious goal echoes the original vision the Siri team had back in 2011, a vision that was sidelined by Steve Jobs in favor of simplicity and due to the limitations of the technology at the time. “This is precisely what was missing from the original Siri,” remarked Dag Kittlaus, the leader of the Siri team during its debut. Kittlaus, along with several important members of his team, later sought to realize this early ambition through a startup named Viv, which has since been incorporated into Samsung's Bixby. Achieving such an intricate system requires widespread developer engagement, a point emphasized during the WWDC, where sessions were dedicated to teaching developers how to integrate their apps with Siri.

Authored by Mark

By Christopher Solomon

Authored by David

Authored by Steven

The advent of Apple's advanced AI technology could be both revolutionary and unsettling, regardless of its seemingly harmless facade. Should Apple successfully realize its ambitions for Siri—enabling it to perform tasks in moments that would otherwise take hours or even days—the repercussions could be widespread and unpredictable. The concern arises: how will Siri decide on our behalf? For instance, when booking flights, might it opt for a less comfortable middle seat for us, deducing our preference for economy, despite our willingness to pay slightly more for an aisle seat? Or, in presenting news, might it inadvertently reinforce our preexisting biases? Regardless of the nomenclature, the introduction of AI by Apple is set to alter our routines significantly. At the Worldwide Developers Conference (WWDC) this week, Apple has committed to being at the forefront of this shift.

Time Leap

In 2016, Apple selected me to announce that its devices were secretly equipped with artificial intelligence. With generative AI becoming increasingly important, the company recognized the necessity of openly emphasizing its commitment to this emerging technology. This focus became the central topic at the 2024 Worldwide Developers Conference (WWDC). However, the challenges related to privacy, oversight, and the composition of training datasets remain largely unchanged from eight years prior.

The impact of machine learning on Apple's offerings is undeniable, yet it remains uncertain if it's transforming the company from within. The principles guiding machine learning appear to conflict with Apple's core values. Known for its meticulous oversight of the user interface, right to the sensors detecting touch gestures, Apple orchestrates every detail and codes with precision. However, machine learning requires engineers to relinquish control, allowing algorithms to evolve and find their own paths to solutions. The question now is whether Apple can embrace the contemporary notion that machine learning can actively contribute to the design process of its products.

Authored by Mark

Authored by Christopher

By David Gilbert

Authored by Steven

"Federighi admits there's quite a bit of internal discussion on the topic. He explains that Apple traditionally prides itself on providing a meticulously planned and controlled user experience. However, when it comes to developing systems that learn from vast amounts of human behavioral data, the outcomes aren't always aligned with what was initially outlined by Apple's designers. Instead, they arise directly from the patterns found within the data."

According to Senior Vice President Phil Schiller, Apple remains steadfast in its approach. "Although these methods certainly influence our design process, our ultimate goal in employing them is to enhance the overall quality of the product we offer."

Inquire About Anything

Fred queries, "Are you convinced that existence persists beyond applications?"

Appreciated, Fred. To dissect your inquiry, it seems you're probing whether there's a possibility of transcending the existing computing model, where an operating system serves as a platform for various applications that facilitate work, entertainment, and social interaction. If such a transformation were to take place, it would mark a monumental shift in the digital landscape. Yet, proponents of generative AI are suggesting a different approach. Rather than launching an application, you would just voice your request to an intelligent assistant and—just like magic—it would be accomplished.

For the proposed scenario to become a reality, numerous changes would need to occur, given that the current digital ecosystem heavily relies on applications. Moreover, artificial intelligence hasn't reached a level where it can perform tasks as efficiently as dedicated apps. However, consider a scenario where Apple's integration with Siri is flawlessly implemented. Imagine instructing Siri with a detailed list of errands, including navigating between locations. Apple could potentially tap into the functionalities of ride-sharing services like Uber or Lyft, allowing Siri to handle all interactions that would typically require the use of their apps. Although you would still be using Uber's service, the necessity for the app itself would be eliminated.

Authored by Mark

Authored by Christopher

Authored by David

Authored by Steven

Currently, apps remain vibrant and active, particularly those enhanced by artificial intelligence.

Send your inquiries to mail@wired.com by including ASK LEVY in the email subject.

Apocalyptic Times Dispatch

In Florida, a state where references to climate change are prohibited in its legislation, new records for rainfall and flooding have been shattered.

Final Highlights

All the innovations Apple introduced at the Worldwide Developers Conference.

Ray Kurzweil isn't taken aback by the current advancements in generative AI; he foresaw these developments years beforehand. Moreover, he claims that human beings have the potential to achieve near-immortality. The ramifications of these assertions are thoroughly explored in our conversation.

Authored by Mark

Authored by Christopher

By [Your Name

Authored by Steven

The harrowing tale of the ill-fated Titan submersible unfolds like a modern-day Titanic disaster.

Has your fancy, top-notch bicycle gone missing? Meet the genius from Mexico who could be behind it.

Ensure you don't skip out on upcoming editions of this column exclusive to subscribers. Get your subscription to WIRED now at a 50% discount for Plaintext readers.

You May Also Be Interested In…

Diving into the Chaos of Programmatic Advertising

What is the number of electric vehicle charging points required in the US to supplant gasoline stations?

A charitable organization aimed to rectify the issues within tech culture but ended up losing command over itself.

Endless Sunshine: Discover the Ultimate Sunglasses for Every Journey

Knight Will

Paresh Dave's

Rogers Reece

Louise Matsakis

Knight Will

Catherine Knibbs

Dave Paresh

Name: Joel Khalili

Additional Content from WIRED

Evaluations and Tutorials

© 2024 Condé Nast. All rights reserved. WIRED may receive a share of revenue from items sold via our website, which is part of our affiliate agreement with retail partners. Content on this website is protected and cannot be copied, shared, distributed, or used in any form without explicit consent from Condé Nast. Advertisement Choices

Choose a global website