Lost in today’s frenzy over a pair of brand-new M3 MacBook Air computers was a not-so-subtle tweak in product language that could signal Apple’s formal entry into the race to build an AI PC.
The press announcement for the new 13-inch and 15-inch MacBook Air ultraportables featuring the latest Apple silicon included two paragraphs saying that the MacBook Air is the “World’s Best Consumer Laptop for AI”. If you didn’t follow the computer industry as closely as I do, you could have dismissed that as some weirdly particular boasting or hyperbole on Apple’s part. I see things a little differently, however.
First, some history. Until 2020, almost all new Apple Macs, including MacBook Airs, had Intel processors. That year, Apple declared its aim to manufacture its own semiconductors and eventually replace all Intel CPUs with its custom system on a chip (SoC), which became known as Apple Silicon. The first such chip, the M1, appeared on the popular MacBook Air M1 (now obsolete).
The march of Silicon
Apple eventually delivered on its pledge, replacing all Intel technology with its own after numerous revisions and upgrades. Intel still controls the majority of the Windows PC market, but in certain ways, Apple is seen as the system pioneer, developing SoCs that are faster and more efficient than anything Intel produces.
Intel’s big plan for countering that perception, and exciting people looking for other ARM-based solutions that can run Windows as fast and efficiently as something like Apple silicon, is to revise its entire chip lineup with Intel Core Ultra processors and, more importantly, the “AI PC.”
The AI component is provided by the Neural Processing Unit (NPU), which will function as an AI coprocessor alongside Intel Core Ultras. Intel has the endorsement of almost all major Windows PC makers, including, probably most importantly, Microsoft. The Redmond software behemoth is currently launching a full-court press against Copilot. The generative AI, formerly Bing AI chat, that it produced with intelligence from OpenAI, appears to be everywhere, and on AI PCs it will appear as a Copilot keyboard button.
What any of us will do with a “AI PC” is uncertain, but we will be discussing these systems throughout the summer and into the Northern Hemisphere’s back-to-school shopping season.
Apple, by some estimates, controls only 17% of the PC market. Even if people believe Apple’s silicon is superior and macOS is a better platform than Windows, they cannot afford to sit back and watch Intel and Microsoft innovate and sell themselves to even higher PC market heights.
We understand AI
This returns us to the “World’s Best Laptop for Consumer AI.” Apple has a point, however. It has been doing AI for a long time, beginning with the addition of its first Neural Engine to the iPhone 8 via the A11 Bionic CPU. This early onboard machine learning technology is directly comparable to the M3’s 16-core Neural engine.
Apple has made no secret of its silicon’s AI capabilities, but it has never put them front and center. That is all changing now.
The corporation has no choice. Apple’s challenge stems from the fact that, unlike Microsoft, OpenAI (ChatGPT), and Google (Gemini), it does not have a generative AI product. Siri is not generative; it cannot generate poems, presentations, or artwork. That has hampered Apple’s attempts to look ahead of the curve.
In the release, Apple specifically mentions Large Language Models (LLMs): “Combined with the unified memory architecture of Apple silicon, MacBook Air can also run optimized AI models, including large language models (LLMs) and diffusion models for image generation locally with great performance.”
The Shape of Things to Come
Running locally, without relying on potentially less secure or slower cloud assistance, has always been Apple’s secret AI sauce. However, Apple is well aware that it cannot win this game unless it allows for cloud-based generative AI.
During demonstrations, I observed the MacBook Air M3 doing both cloud-based Microsoft CopIlot prompts and local generative tasks using programs like as Luminar Neo, which can take a fuzzy midnight shot and add generative information to make the image useful. In all cases, their performance appeared almost instantaneous and was easily comparable to that of a cloud-based generative AI.
The objective of demonstrating these apps and making these announcements, however, is not simply to inform the world that Apple also does Gen AI. I believe it is preparing us for what is to come.
It is not only about new products and press releases. Apple CEO Tim Cook currently takes practically every opportunity to make huge promises about Generative AI (remember when he hyped “AR”? What a difference a letter makes.
Cook understands that Apple technology is more than ready for Large Language Models and Generative AI for images and text, and we’ll see Apple take advantage of all that power beginning with WWDC 2024 in June.
That’s Apple’s message about AI: You haven’t seen nothing yet.
