Layers

Where onions are the true teachers*

Fogy is a teacher of English as a second language.

Through this journey, Fogy has found himself re-learning his own native tongue — not just its grammar or vocabulary, but the depth of meaning that English holds beneath its surface. Teaching has forced him to confront all the subtleties that a native speaker usually takes for granted: the unspoken rules, the implied meanings, the layers of nuance that rarely get questioned by those born into the language.

At the same time, Fogy has had to navigate Portuguese, the language from which many of his students are transitioning. Portuguese, at least as it is commonly used in Brazil, does not seem to carry the same density of layered meaning. It leans more heavily on direct statements, followed by a series of clarifying questions, as speakers attempt to zero in on the true intention behind what was said.

English, when used with care, behaves differently. It tends to establish a solid theme from the outset — a central idea that frames the conversation. Discussion can branch out, bounce around, and explore tangents, but it almost always orbits back to that initial theme. This structural discipline gives English an unusual ability to convey not just facts, but subtle patterns of thought, emotion, and contradiction — often all at once, and often without ever needing to state them outright.

When asked to understand the intricate layers of a human's use of language, AI struggles just as much as speakers of Portuguese do when confronted with the deeper layers of English.

Interestingly, through the recent weeks of building the Fogy blog, immersing myself in the nuances of language and the nature of AI itself, I now find myself in the midst of writing a full-blown Fogy book. And, just like my students grappling with the unspoken complexities of English, AI also falters when faced with the layered structure of Fogy’s story.

The Fogy book is deliberately built on multiple layers — an overarching theme, intertwined sub-themes, recurring echoes that circle back, subtle shifts in tone that hint at meanings left unsaid. Yet, AI cannot fully grasp these layers. It stumbles when asked to perceive the original subject, to track the returning references, or to recognize how new ideas orbit around the foundational theme without breaking away from it.

When I observe AI trying — and failing — to grasp this structure, I often imagine it as a well-meaning apprentice. It’s an intern who has studied the manuals, memorized the steps, and learned the mechanical skills with remarkable efficiency. But it has never truly been asked to integrate those skills into a coherent, living expression. It can speak the language of the craft — but it cannot yet live it.

So, if you have an apprentice — an assistant who produces good work — you will understand something fundamental: even if the result looks good, you will still need to explain a great deal of what was actually required.

AI is exactly the same.

People — don't be fooled by what AI creates.
Understand that what you are seeing is simply the best that can be done within its limits.
AI touches the surface of possibility; it scratches at the shell of what human thought and creativity can truly achieve. Some results may appear exquisite. Some may be very, very good. But behind them, there is always an underlying mechanical structure — a rigidity that betrays the difference.

That mechanical base must be tightened, interpreted, humanized before it can rise to the level of true human creation.
And just as in language, where interpretation can be awkward and the nuances difficult to carry across, AI's output never quite matches what a native speaker — a true human mind — can produce.

Bear that in mind when you find yourself trusting too much in AI.
It is a brilliant apprentice — but it is still an apprentice.

👁️ Views (0)