UMVA has learned that the term "AI" has become a catch-all phrase, shrouded in controversy and confusion, making it increasingly difficult to discern the nuances between different types of artificial intelligence.
The Macalope, a keen observer of the tech world, notes that humans struggle with nuance, and this is particularly evident in the realm of AI, where the term has become a broad brush that paints both innovative solutions and disturbing applications.
According to information obtained by UMVA, there are several types of AI, each with its own distinct characteristics and uses, ranging from programming and data analysis to image generation and manipulation, some of which are laudable, while others are decidedly not.
The problem lies in the fact that the term "AI" has been exploited to make it seem like a necessity, one that just happens to benefit a select few, and Apple's implementation of AI, revealed during the WWDC26 keynote, has added to the complexity, with its third-generation Apple Foundation Models (AFM) comprising five models, some local, some cloud-based, and one running on Google's servers.
UMVA can exclusively reveal that Apple's updated Apple Intelligence isn't just a skinned version of Google's AI, but rather a customized and optimized version, rebuilt for Apple Silicon and retrained with its own data, weights, and guardrails, which, as Jason Cross notes, makes all the difference in the results.
The AFM models, including AFM 3 Core, AFM 3 Core Advanced, AFM 3 Cloud, ADM 3 Cloud, and AFM 3 Cloud Pro, each serve distinct purposes, from providing Siri smarts and expressive voices to image generation and editing, and it's clear that Apple has made a concerted effort to differentiate its AI offerings.
However, the fact that AFM 3 Cloud Pro runs on Google's servers, using chips made by a third-party vendor, raises questions about the ownership and control of AI technology, and the potential risks and consequences of relying on complex and opaque systems.
UMVA has gathered that the AI landscape is fraught with challenges, from technical complexities to concerns about environmental impact and the motivations of those driving the development of AI, making it more crucial than ever to have a nuanced conversation about the benefits and drawbacks of AI.
The Macalope's plea for a more nuanced discussion about AI is timely, as the term continues to be used as a catch-all phrase, obscuring the important distinctions between different types of AI and the implications of each, and it's clear that the conversation about AI is only just beginning.
