Apple Shifts Gears: From Cars to Cutting-Edge AI

Apple has relinquished one of its most ambitious endeavors, the pursuit of automotive engineering, and is contemplating relocating 2,000 employees to its artificial intelligence enclave. Informants closely associated with Apple disclosed to the author that personnel from both the Vision Pro and automotive project cadres have previously transitioned to artificial intelligence undertakings.

In the eloquent words of Steve Jobs: “Determining what ventures to abstain from is as crucial as discerning those to pursue.”

Despite a decade-long endeavor to manifest automotive innovation, amidst the formidable presence of Think Different’s OpenAI, Apple has at last roused from its reverie and resolved to fully engage in the AI arena.

The global narrative suggests that Apple’s complete pivot towards AI is long overdue, yet Apple merely adheres to the dictum of “early to bed, early to rise.” From Steve Jobs’s embryonic forays into Atari’s “Brick Breaker” game to Siri, which delivered a masterclass to all voice assistants in 2016, Apple has inscribed several momentous chapters in the annals of AI.

At present, Apple can no longer abide the duopoly dominance of OpenAI and Google and has opted to redouble its investment in AI. As the sole tech entity potentially possessing full-stack capabilities spanning the strata of computational potency, intermediate frameworks, model architectures, and application interfaces, will Apple’s foray herald groundbreaking strides in the AI domain? What distinct advantages does the tardy arrival of Apple proffer in this global AI skirmish?

Surveying the global AI landscape, Microsoft marshals OpenAI to besiege rivals, yet finds itself besieged; Google pursues OpenAI, but merely treads in its wake; domestic manufacturers endeavor to navigate the waters alongside OpenAI, yet possess innate advantages. The configuration of this tripartite world remains indeterminate. The vast expanse holds much promise; with time and effort, hegemony may be attained, and Apple may flourish.

The Early Evangelist of AI

According to data from the IIF, global indebtedness is poised to reach a record zenith in 2023, cresting at $313 trillion. Amongst the burgeoning liabilities,

In the realm of AI, how premature were Apple’s initial strides? Tracing its genesis reveals an inception predating that of any extant AI juggernaut.

In 1975, whilst the illustrious saga of Gang Leader Qiao was yet unfinished and Apple remained inchoate, a youthful Jobs was engrossed in honing his craft and battling digital adversaries at Atari Game Company. Atari held sway over the gaming realm akin to Nintendo in the late ’80s and PlayStation in the ’90s. Jobs, then a callow twenty-year-old, alongside Steve Wozniak, labored tirelessly for four nights to engineer the hardware design of the game “Arkanoid.”

“Arkanoid” garnered instant acclaim upon its release, ascending to legendary status in the pantheon of game design and subsequently serving as the substrate for the DeepMind team’s deep learning endeavors.

Initiating from the chess-playing algorithms devised by Turing and Shannon, the progenitors of AI, the nascent discipline of artificial intelligence aspired to fashion a program capable of besting human adversaries in gameplay.

Thus, DeepMind amalgamated reinforcement learning with deep neural networks to engender a system proficient in mastering Atari games. Following myriad iterations, the average score attained by the deep learning network in “Arkanoid” surpassed that of humans by tenfold. DeepMind garnered renown, culminating in its acquisition by Google a year hence, heralding the advent of AlphaGO and inaugurating the latest wave of AI advancement.

In the domain of natural language processing and speech recognition, pivotal milestones in AI evolution, Apple once outpaced every mobile phone conglomerate.

In 2010, Apple procured Siri. Subsequently, in 2016, Apple unveiled the sentient voice assistant Siri on the iPhone, exhibiting a quantum leap in performance vis-à-vis its counterparts. “Occasionally, an enhancement in performance is so pronounced that it prompts reevaluation to ensure no detail eludes scrutiny. The debut of Siri exemplifies such an occurrence,” remarked Apple engineers.

Missed Opportunities Amidst the AI Renaissance?

As per the IIF’s data, global indebtedness is slated to burgeon to a staggering $313 trillion by 2023. Within this burgeoning indebtedness,

Although Apple indirectly catalyzed the preceding AI renaissance, it finds itself caught off guard amidst the nascent AI epoch precipitated by OpenAI. Despite its concerted efforts, Apple has remained circumspect and understated.

In this nascent AI epoch, Apple’s endeavors encompass a trifecta, none of which are without their imperfections.

Firstly, through the acquisition of myriad AI startups. According to the latest report from market research agency Stocklytics, Apple procured a total of 32 AI enterprises in 2023, integrating the acquired AI technologies to enhance product offerings. For instance, in 2020, Apple acquired Voysis, a Dublin-based purveyor of voice AI technology, augmenting the conversational prowess of Siri.

Secondly, the development of expansive models and AI conversational agents. In July 2023, revelations surfaced regarding Apple’s endeavors towards the creation of a large-scale model dubbed Ajax, alongside an internal chatbot christened Apple GPT. While Ajax leverages Google’s machine learning framework, Google Jax, as its foundational scaffold, it predominantly serves as a driving force for internal endeavors, remaining inaccessible to consumers. Notwithstanding, some Apple personnel contend that it essentially mirrors existing entities such as Bard, ChatGPT, and Bing AI, devoid of any novel features or technological innovation.

Thirdly, Siri is speculated to be closely entwined with the large-scale model. In 2018, Giannandrea, erstwhile head of Google AI, assumed leadership of Apple’s artificial intelligence and machine learning division, concurrently assuming stewardship of Siri. Siri has encountered censure for stagnation, with Siri co-founder Dag Kittlaus intimating that Siri may not have realized its full potential post-acquisition by Apple. Following the paradigm shift catalyzed by ChatGPT, Siri finds itself under heightened scrutiny.

Evidently, although Apple has devoted copious time and resources to AI, its endeavors have been marked by hesitance. In this regard, Cook has tactfully articulated that Apple will progressively integrate artificial intelligence into its product portfolio, albeit “after thoughtful deliberation.”

Previously, when queried regarding Apple’s endeavors in generative AI, Cook’s rejoinder was a cryptic “stay tuned.” “Apple has yet to make a resounding impact in the realm of AI,” remarked Brian Mulberry, an investment manager representing Apple shareholders. Amidst the current AI renaissance, Apple’s ardor for AI pales in comparison to the fervor exhibited by Microsoft, Google, and OpenAI.

Apple’s AI Odyssey

Subsequent to the revelation of its departure from automotive pursuits in favor of AI, Cook, at the shareholder meeting on February 29, adopted a markedly different stance, affirming the company’s intent to “chart new frontiers” in the realm of generative artificial intelligence in 2024. He emphasized: “We believe this will engender transformative opportunities for users.”

Hence, what are the merits of Apple’s tardy entry? Will it, as Cook asserts, blaze new trails in the AI realm?

Qiu Chen, overseas partner at Huaying Capital, posits that Apple’s advantage lies in its potential status as the solitary technology behemoth possessing comprehensive capabilities across the computational, intermediate, model, and application strata. The confluence of proprietary artificial intelligence chips, cloud computing infrastructure, and algorithmic optimization yields substantial enhancements in model system optimization.

Consequently, buoyed by its substantial investments in AI, Apple, leveraging its unique advantages, is poised to carve a bespoke path in the AI landscape and potentially emerge as a preeminent force amidst the intense AI milieu.

Subsequently, let us examine Apple’s prowess across each of the aforementioned strata.

Computational Potency Stratum: Presently, the training aspect of AI leans heavily on NVIDIA. Apple, endowed with its own data centers spanning China, Europe, and the United States, has amassed a formidable arsenal of NVIDIA GPUs, thus averting the specter of computational scarcity. On the inference front, Apple’s sustained investments in chip design R&D and comprehensive toolchains render it fully capable of deploying inference acceleration solutions predicated on proprietary chipsets, whilst accommodating large-scale proprietary models. Already, Google TPU and startup Groq have achieved commensurate results vis-à-vis Nvidia in this domain.

Data Stratum: The iOS ecosystem, boasting an extensive user base, furnishes Apple with an abundant corpus of data and user behavior. Since its integration into the iPhone 4S’s Siri in 2011, it has amassed a decade-spanning trove of user conversational data.

Application Stratum: The extant iOS ecosystem affords a platform and audience for content generation services encompassing text, imagery, audio, and video, catering to end-users.

Qiu Chen contends that commencing from the aforementioned large-scale models and proprietary chipsets, in tandem with a synthesis of software and hardware endeavors, Apple is poised to address the dual facets prioritized by OpenAI and NVIDIA. Although the attendant costs remain formidable, Apple is poised to reallocate resources following the recalibration of its business priorities, with the probable sequence being: MR (Mixed Reality) → large-scale models → autonomous vehicles → embodied intelligence.

Endowed with the aforementioned advantages, in what domains will Apple, now fully committed to AI, engender transformative opportunities for users?

Let us commence with the burgeoning domain of AI-enabled mobile devices. While domestic competitors made inroads into large-scale edge models a year prior, Apple’s sustained investments in chipsets and indigenous R&D confer upon it superior operational efficacy vis-à-vis rival mobile phone manufacturers. Leveraging its prodigious computational resources encompassing CPU, GPU, and NPU, Apple is uniquely positioned to surmount the exacting demands of running large-scale models on devices, characterized by stringent requisites pertaining to memory speed and resource allocation. Nevertheless, true to Apple’s precedent, it is likely to eschew overt references to AI-infused smartphones and edge models, instead focusing on application scenarios. What novel solutions will Apple unveil, distinct from those proffered by its counterparts? The forthcoming releases of iOS18 and iPhone16 are poised to unveil the answer.

For Vision Pro, Apple’s vanguard mobile terminal platform, AIGC (Apple’s Interactive Graphics Chip) promises substantial cost efficiencies and efficacy improvements in 3D scene construction. Real-world data and user interactions gleaned through onboard sensors and cameras can be seamlessly integrated with AIGC for robust big data analytics and modeling, engendering superior real-time interaction and user experience feedback.

Insiders intimate that Apple’s cessation of automotive pursuits is likely a transient postponement of its automotive ambitions, awaiting breakthroughs in autonomous driving technology before unveiling more mature autonomous vehicle offerings. Sora’s acumen in discerning and navigating the physical laws governing the real world, alongside its capacity to anticipate object trajectories, augur novel paradigms in autonomous driving, rendering AI the sine qua non. Thus, Apple’s pivot away from automotive pursuits towards AI is not tantamount to an abandonment of automotive aspirations but a recognition of the indispensability of AI in realizing autonomous driving ambitions.

error: Content is protected !!