A friend of mine Harry Hoffman recently shared an article by Sangeet Paul Choudary titled “The many fallacies of ‘AI won’t take your job, but someone using AI will’”. The article offers a thought-provoking analysis of how AI is transforming work. While the article covers multiple professions and systems, I’d like to focus on what this means specifically for software and web developers and how those professions are experiencing a transformation similar to what typists and other roles mentioned in the article faced.

I know that the mere writing of this article is going to strain some of my relationships. I am not taking this lightly, I love coding, especially the late night coding session, but this is something I believe is happening and I want us to start talking about it. I have had many arguments about how AI will impact our profession, but many times it is dismissed, because “the context windows are too small” or “there are too many security implications”, etc. But one quote always comes to mind “AI is the worst it will be, today.” This has proven to be true time and time again. Today’s models, platforms, agents, MCPs, are being iterated at speeds we have never seen before. MCPs to my knowledge were not even around when AI was first released to the general public.

Let’s use web developer transitions as an example:

1. Early days: Manual HTML coding, custom CMSs
2. AJAX revolution (~2005): Dynamic content without page reloads transformed user experiences
3. Mobile/responsive design era (~2010): Forced developers to think beyond desktop
4. Middle era: CSS/JS frameworks emerged, PHP/.NET for CMS development
5. JavaScript frameworks (Angular, React, Vue): Single-page applications changed architecture
6. API-first/headless CMS approach: Decoupled frontend and backend development
7. Current trend: No-code tools empowering marketers

Each transition required developers to adapt their skills or risk obsolescence. The profession hasn’t disappeared—it’s continuously transformed as lower-level tasks become commoditized while new specialties emerge at higher levels of abstraction.

We've been through this before: AJAX, mobile, frameworks, APIs. Each time, the job evolved. AI coding isn't different—it's just happening faster.

The Typist Parallel

The most striking parallel from the article is the example of typists in the late twentieth century. As Choudary notes, when word processors emerged, typists believed that “Word processors won’t take your job, but someone using a word processor will.” However, what actually happened was more profound:

“Typists weren’t outcompeted by better typists. They were displaced by a new system design in which typing no longer justified a full-time role. […] Typing became embedded across all workflows. From a specialized task requiring specialized skills, it became a basic task that everyone could perform.”

Today’s software developers face a similar inflection point. The emergence of AI coding assistants like GitHub Copilot, Claude Code, Augment, and other LLM-based tools has begun to transform the nature of coding itself. Just as word processors made document editing cheap and accessible to everyone, these AI tools are making certain aspects of coding more accessible to non-specialists.

Beyond Task-Level Thinking

Focusing on automation versus augmentation at the task level misses the bigger picture—how entire systems are being restructured. For developers, this means that the value isn’t just in writing code anymore, but in understanding how to design systems that leverage AI effectively.

The fallacy of task-based thinking is particularly relevant here. While we might believe that AI will either automate coding (replacing developers) or augment coding (making developers more productive), the reality is that the entire system of software development is changing. The value of certain coding tasks is diminishing while other capabilities are becoming more important.

Skills Commoditization

Another key insight from the article is how certain skills become commoditized:

“Companies that adopt AI for task acceleration will soon realize that when tools are widely available and easily replicated, productivity becomes a commodity.”

Basic coding skills are becoming commoditized through AI. The ability to implement standard algorithms or create boilerplate code is increasingly handled by AI tools. This commoditization is reshaping what makes a developer valuable.

When everyone has AI that can write React components, the value isn't in writing React—it's in knowing what to build and why.

From Coder to System Designer

Just as the article describes shifts in various fields like basketball (from fixed positions to fluid roles), developers are experiencing a rebundling of their roles. The job is evolving from writing code to designing systems and workflows where humans and AI work together effectively.

This mirrors what Choudary calls “the static jobs fallacy”—the mistaken belief that jobs remain fixed units while tools change. Instead, entire roles get unbundled and rebundled around new priorities. For developers, this means less focus on implementation details and more emphasis on architecture, coordination, and problem framing.

The Power Shift

The article discusses how tools like Excel shifted organizational power toward those who could control spreadsheet cells. Similarly, AI coding tools are shifting power within development teams. Those who can effectively prompt, direct, and integrate with AI tools may gain influence, while those who primarily contributed through writing code may see their relative value decrease. Imagine you spent all your time training to be a coder only to find that the majority of your time is doing AI code reviews.

What This Means for Developers of the Future

Focus on system design as much as implementation: As AI handles more implementation details, value shifts to those who can design effective systems and architectures.

Develop prompt engineering skills: Understanding how to effectively direct AI tools becomes a crucial skill, similar to how Excel proficiency became valuable in the 1990s.

Strengthen domain knowledge: Deep understanding of business domains becomes more valuable as coding itself becomes commoditized.

Embrace coordination roles: As development becomes more modular with AI handling components, the ability to coordinate and integrate becomes critical.

Prepare for salary pressure: As the article suggests with session musicians, we may see more work but potentially at lower rates as basic development skills become easily accessible.

The developers who thrive won't be the ones who code fastest—they'll be the ones who understand what should be built and can orchestrate AI to build it.

The transformative impact of AI on software development isn’t simply about whether AI will replace developers or make them more productive. Rather, it’s about how the entire system of software development is being restructured. The most successful developers won’t be those who simply use AI tools to code faster, but those who recognize how the field is being fundamentally reshaped and position themselves accordingly.

As Choudary eloquently puts it: “The real advantage is not in making existing workflows faster, but in being first to build the new ones that won’t need those steps at all.” For developers, this means looking beyond code production to the new forms of value creation that are emerging because of AI.

The typist didn’t disappear because they weren’t good at typing. They disappeared because typing became something everyone did as part of their job. We might be witnessing the same transformation with coding—where the act of writing code becomes a basic skill embedded in many roles, rather than a specialized profession unto itself.

The question isn’t whether you’ll still have a job as a developer. The question is: what kind of developer will you become when coding is no longer the primary source of your value? The answer to that question will determine not just your career trajectory, but your relevance in an AI-integrated world.

This transformation is happening now, not in some distant future. The developers who recognize this shift and begin adapting today will be the ones who shape the new landscape of software development. Those who cling to the old model of value creation may find themselves in the same position as typists who insisted that word processors would never replace their specialized skills.

The choice is ours to make, but we need to make it soon.

Vatché

Tinker, Thinker, AI Builder. Writing helps me formulate my thoughts and opinions on various topics. This blog's focus is AI and emerging tech, but may stray from time to time into philosophy and ethics.