Differences between Apple’s thinking about the cost and deployment of AI compared to the big tech competitors

Apple’s new Vision Pro virtual reality headset will be shown at Apple’s Worldwide Developers Conference (WWDC) at the Apple Park campus in Cupertino, California, on June 5, 2023.

Josh Adelson AFP | Getty Images

Since 2015, Apple has acquired more than two dozen AI companies. They are hardly household names, including Emotient, Laserlike, Drive.ai, AI.Music and WaveOne. But Apple’s engineers have steadily upgraded the technology it bought into smartphones, computers and watches, music and TV streaming services, countless mobile operating systems and applications, as well as the Vision Pro mixed-reality headset, due next year.

Cupertino-based Apple doesn’t talk publicly about AI acquisitions and is generally tight-lipped about its overall strategy in the space, including long-term in-house R&D in AI, even as big tech rivals Microsoft, Google, Meta and Amazon are talking about this. They advertise their AI chatbots and Large Language Model (LLM) platforms.

“It’s the DNA of Cook and Cupertino. They tend not to talk until it’s published,” said Don Ives, an analyst at Wedbush Securities.

Advertising isn’t the only difference that can be gleaned from Apple’s AI approach to date and its future implications for its consumer-centric business model.

While its competitors are focused on building autonomous generative AI models, Apple is targeting machine learning infrastructure.

“Apple is looking at acquiring teams of leading talent in each field that can apply machine learning techniques to specific consumer products,” said Brendan Burke, emerging technology analyst at research firm PitchBook, which has tracked Apple’s 30 AI acquisitions worldwide. bring. For the past eight years, he said, “This has led to an acquisition strategy focused primarily on consumer applications of AI, but also operational techniques for deploying machine learning and edge devices, as well as limited bets on the future of deep learning and more horizontal technologies.”

Apple, which is scheduled to report earnings on Thursday, did not respond to requests for comment.

Apple has been more cautious when talking about its AI ambitions, whether when CEO Tim Cook appeared on “Good Morning America” ​​or during an earnings call with analysts and investors in August. “We see AI and machine learning as core and fundamental technologies that are integral to almost every product we make,” Cook said on the call. “And of course, we’re doing research in a wide range of AI technologies, including Artificially productive, for years,” he said. “We will continue to invest, innovate, and responsibly advance our products with these technologies, with the goal of enriching people’s lives. We tend to announce things that come to market, and that’s our goal, and I I want to stick to it. To it.”

Until then, the latest iPhone models and watches showcase the power of Apple’s artificial intelligence. The iPhone 15 features several artificial intelligence technologies, such as Personal Voice, which allows users to synthesize a voice similar to their own voice to speak the words they type in FaceTime and phone calls, and Live Voicemail, a real-time transcription of messages. The upgraded camera uses machine learning artificial intelligence to distinguish between a person and an animal in the frame. The Apple Watch Series 9 and Ultra 2 incorporate artificial intelligence into the double-tap feature for easier task completion, a brighter display, smarter Siri, and advanced health monitoring.

In announcing the headset, which it calls a space computer, Apple said artificial intelligence will also be built into Vision Pro, specifically its integration with FaceTime. Users can create avatars of themselves through machine learning techniques, allowing them to do things together like watch movies, browse photos, or collaborate on a presentation.

However, app-by-app and device-by-device advancements haven’t shielded Apple from criticism of being late to the game.

Laura Martin, a tech analyst at Needham, is among those who say Apple is “way behind” its big tech rivals. “The future of artificial intelligence is generative, and Apple is not doing that,” he said. However, Martin added, “I’m not sure it matters because they have an ecosystem, an economic model” geared towards machine learning. “Apple’s ecosystem is good and will continue to grow,” he said.

As shown in Microsoft’s recent earnings report, an early lead in generative AI is paying off for competitors. But one advantage Apple has that its tech rivals don’t have: the highest installed revenue in the world, more than two billion users running its iOS operating software on devices.

That’s why bullish analysts don’t see Apple competing directly against AI platforms pursued by Microsoft, Google and Amazon. Comparisons today to the first wave of cloud adoption and generative AI ecosystems, now referred to by some as Cloud 2.0, are a good place to look at why this might be less of a race than a parallel development process. Even as Amazon Web Services, Microsoft Azure, and to some extent Google Cloud gained market share as more companies migrated to cloud-based computing, Apple made no play for that market.

“They’re not looking at AI from the outside. They’re going to be a big player in AI even if they’re perceived as late to the game,” Ives said. “We believe they are on the verge of introducing an AI App Store next year,” he said.

Billions of dollars spent on artificial intelligence

Last week, Bloomberg reported that it will spend $1 billion a year on generative artificial intelligence in a race to “catch up” with Apple. Ives says he believes that’s what Apple is spending on AI, and has already spent. He expects Apple to spend $5 billion a year on AI, and says it has spent $10 billion on AI research and development over the past few years to lay the foundation for developers to build AI applications at a level that is feasible. Make a comparison with Microsoft. “This is the holy grail for them, the AI ​​App Store. This is where they can monetize in a historic way,” Ives said.

It’s a growth opportunity that touches on one of the most important aspects of Apple’s story in recent years, the growth of its services business, which alone now has a $100 billion revenue stream. In an AI App Store, the opportunities could extend far beyond Apple’s core offerings like Music, to a wide range of health and fitness apps, for example, and ultimately $5-10 billion in additional revenue per year from the service. Count Ives

Cook’s mere mention of generative AI to analysts has fueled speculation about what Apple will or won’t do in the highly competitive field, leading to Microsoft’s partnership with OpenAI and ChatGPT and Bing’s AI chatbot, Bot. Google’s Bard and LLM meta. Meta artificial intelligence based virtual assistant.

In July, Bloomberg reported that Apple is working on AI tools to challenge these bots with its LLM, Ajax, and has created a chatbot called Apple GPT for internal use.

In August, the Financial Times reported that Apple was ramping up its AI expertise by hiring dozens of people in California, Seattle, Paris and Beijing to work on LLMs. In fact, Apple’s career website invites job seekers to “join a team of researchers and engineers with a proven track record in a variety of machine learning methods, including artificial intelligence models.

However, there is a downside to its ecosystem, which adheres to data privacy and design excellence, and may be an obstacle to anything Apple does in generative AI, or at least explain what it’s doing, Burke said. Why might it start up slower? “Apple only wants to deliver products that are secure, privacy-focused, and meet internal quality standards, so that limits the speed at which they can be deployed,” he said. “So I think Apple will wait to release any platform technology until it really adds to its ecosystem.”

Meanwhile, Burke said, it’s reasonable to assume that Apple will continue to invest in AI, whether by buying more startups or partnering with developers, and that it won’t go unnoticed by deals made by rivals. Referring to Amazon’s recent $4 billion investment in Anthropic, a startup that launched an artificial intelligence chatbot called Claude 2, he said, “Apple could make a similar push to partner with leading startups to complement its efforts.” feel its internal research and development.” “Especially as startups continue to achieve breakthrough results with LLM and incorporate new image- and video-based features that could be relevant for Apple. It could be a valid complement to the work that he is doing it.”