Adapted from a post I wrote on my Instagram story in February 2026.

When I first watched Culinary Class Wars, I found Chef Ahn Sung-jae’s persona genuinely unusual. What caught my attention was his vocabulary. Again and again he reaches for phrases like “the vegetables are cooked evenly,” “it lacks a kick,” or “the harmony of textures.” These are tools for slicing the physical state of a dish and the finished quality of its flavor into precise units of judgment — a way of revealing a resolution that only an expert possesses. Where a layperson lumps everything together as simply “tasty” or “undercooked,” the expert uses precise language to segment a phenomenon and give it structure. The way Chef Ahn judged the dishes is a fine, almost emblematic example of how expertise expresses itself through language.

The same pattern shows up in other fields. When a designer says they’re “cutting out the subject” or that “the blacks are crushed,” they are using a high-efficiency compression technique to convey, in a single phrase, a complex physical state — the separation of an image, or the loss of detail in the shadows. It’s the same reason scientists borrow the language of mathematics to describe the world. Instead of stretching a phenomenon out in long strings of everyday words, they express what they mean through rigorously defined conceptual terms.

In the past, expertise rested on the proficiency to physically perform a particular skill. Coding ability, drafting skill, calculation speed — these were the measures of an expert. But now that artificial intelligence is taking over the domain of execution, the essence of expertise is shifting from performance to definition. Expertise now means acquiring the systematized language of a field and, through it, being able to frame the essence of a problem.

Ludwig Wittgenstein declared that “the limits of my language mean the limits of my world.” This insight holds real meaning in the context of modern expertise as well. The specialized vocabulary of any field has the character of a data protocol that compresses complex concepts to a high degree, and the expert uses this language to observe phenomena at a finer resolution. Learning the language of a particular field is not merely an act of memorizing new words; it is closer to a process of expanding the horizon from which you perceive the world, through that field’s own logical system. In the absence of such a linguistic foundation, you may well possess a high-performance engine like AI and still find yourself constrained in drawing out its full potential. There is a sense in which the precision of the language a user commands determines just how concrete a destination the AI can reach.

In this process, the point that distinguishes the expert from the charlatan often reveals itself in their attitude toward language. The charlatan tends to use the shell of technical jargon as a device for projecting authority or muddying the point. They reel off ornate rhetoric, but remain vague about the logical cause and effect, or the actual principles of implementation, that lie behind that language. The true expert, by contrast, wields language as a tool for communication and problem-solving. They can verify the validity of what the AI produces, reinterpret complex terms to suit the situation, or translate them into ordinary language. In other words, the key difference is that they don’t stop at merely using or stringing together language — they have a firm grip on the logic that the language contains.

As the interface between humans and technology converges around natural language, paradoxically, fluency in the language of a specific field becomes an even more meaningful competitive edge. The expert sets the direction of the AI’s execution by clarifying the “what” and the “why.” Here, specialized vocabulary serves as a specification that reduces the margin of error in communicating with the AI, and acts as a primary verification mechanism for judging whether the output matches the intent. The more technical execution is automated, the more the human role concentrates on the domain of design — projecting intent precisely through language.

To have expertise is to acquire the language of a field and, in doing so, to gain a perspective that reconstructs the order of things. The more direct execution recedes in an age like this, the more it matters how we name and structure phenomena. Language becomes more than a mere means of communication; it becomes the essential medium through which we convert our own intent into technical reality.

So in the end… the limits of the language we command turn out to be the limits of the world we can build through AI.