Coding's Future: Will It Disappear or Become 'The Way'?

John Carmack, the developer behind the game Doom (DOOM), has drawn an intriguing analogy: as AI generates code, direct coding by humans may one day evolve like Japanese martial arts. At first glance, it might sound like he's predicting coding's demise. But Carmack's core argument isn't about disappearance—it's about a shift in coding's role and value.

He draws inspiration from Miyamoto Musashi's 'The Book of Five Rings,' comparing how martial arts transformed from survival skills on the battlefield to disciplines practiced for personal growth today. Similarly, early programmers had to work at machine-level code, but over time, abstraction through compilers, high-level languages, and libraries reduced the need for direct coding.

Now, generative AI is accelerating this trend with 'vibe coding,' where developers describe desired functionality in natural language rather than writing code line by line. Instead of translating requests into programming syntax, they converse with AI to refine outputs: 'Make a login feature,' 'Display data as a graph,' or 'Adjust for mobile screens.'

Carmack's martial arts analogy holds: just as swordsmanship isn't about how fast you swing but when and where to strike, coding in the AI era will focus less on writing code and more on identifying problems and defining solutions. The ability to spot user needs and connect them with technology becomes increasingly important.

As AI lowers the cost of generating code, the key differentiator shifts from technical execution to problem discovery and ideation. Previously, turning an idea into a product required significant time and resources. Now, simple prototypes can be built quickly by anyone with an idea—meaning that what to build becomes more critical than how to write it.

Developers won't vanish; instead, their role will evolve from implementing requirements to discovering problems and using AI to rapidly turn ideas into reality. Vibe coding exemplifies this shift: the focus isn't on the code itself but on clearly defining what needs to be built—why a button is needed, where it should appear, and what happens when clicked.

Just as modern martial arts practice has separated from battlefield use while retaining its value, hand-coding may transition from an essential production task into a refined skill. Understanding how computers work and designing software structures will remain relevant but no longer be the only path to creating software.

Carmack's insight isn't about coding disappearing—it's about it becoming 'the way' (a martial arts concept), where the focus shifts from writing code to knowing what to build, why, and for whom. As AI makes code production cheaper, the value of identifying new intersections between people and technology will grow.