Content is Code: How AI is Shifting Creation Left

In a world where content creation is constantly evolving, the way we communicate technical information is undergoing a significant transformation. Matt Palmer, Head of Developer Experience at Conductor, argues that code is rapidly becoming the new language of content, marking a fundamental shift in how technical communication is approached.

Palmer, who previously led DevRel at Replit, presented his insights on how content is moving 'left' to code as the source of truth. He breaks down this evolution into three distinct eras: the handcrafted era, the era of expensive code, and the current era of cheap code, largely thanks to AI.

Era 1: The Handcrafted Era

In the earliest stage, content creation was entirely manual. The only levers available were time, money, and skill. If a person or company needed something created, they either had to do it themselves or hire an expert, such as an agency or a professional motion graphics designer. This era was characterized by significant investment in human capital and time for each piece of content.

Era 2: Expensive Code

As code and cloud infrastructure matured, they became more accessible, but also more complex. This led to an era where code was a viable option for content creation, but it was expensive. Professional software engineers were the gatekeepers, and their time was primarily dedicated to core engineering tasks rather than content production. The cost of hiring these engineers, or agencies with similar expertise, meant that content creation, even for things like websites or documentation, was a significant investment and often not the primary focus for engineering teams.

Era 3: Cheap Code and the Rise of AI

We are now in an era where code is cheap, largely due to the widespread availability of AI tools. Palmer highlights that AI, such as models from Anthropic and OpenAI, can now be used to generate videos, documentation, website designs, and motion graphics. This has democratized the creation process, making it faster and more accessible than ever before.

Palmer shared his personal journey, noting that his preferred medium for creating assets for his presentations is now TypeScript, which he admits he's still learning. He emphasizes that the fastest way to build any asset, from videos and websites to slides and images, is through code or code generation.

The True Cost: Structure and Conscientiousness

While code generation is becoming inexpensive, Palmer posits that the truly expensive elements are structure and conscientiousness. He argues that maintaining a structured codebase, adhering to brand guidelines, managing design tokens, separating concerns, and producing clean, well-documented code are the real challenges. These factors are crucial for differentiating between generic, unpolished output and professional, high-quality assets.

"Structure is the difference between AI purple gradient slop, and something that looks professional and polished," Palmer stated. He further defined conscientiousness as "the quality of being meticulous, careful, and guided by a strong sense of moral or professional duty." AI, he believes, rewards these traits, not necessarily the highest technical skill.

The Future of Content Engineering

Palmer predicts that the most successful teams in the coming years will be those that instill discipline and rigor in their software development processes, enabling them to effectively shift content creation left towards code. He forecasts that 2026 will be remembered as the year of the creative technologist in the DevRel space, while 2027 will be the year of the content engineer.

Looking ahead, Palmer anticipates the rise of declarative and robust content pipelines capable of automating the creation of content walkthroughs, documentation, screenshots, and product updates, tasks that were once manual and time-consuming. These advancements, powered by code and AI, will fundamentally change how technical communication is produced and consumed.