Unprompted: The End of Coding?

Unprompted is an occasional opinion column from Kunal Gupta for Pivot 5 readers

Last week, I found myself in a somewhat unexpected position. Armed with nothing but my voice and a clear idea of what I wanted to build, I watched as Lovable's AI turned my spoken descriptions into a fully functional web application. No coding environments, no syntax errors, no Stack Overflow tabs frantically opened in desperation.

Twenty years ago, I graduated with a software engineering degree. For the following fifteen years, I built and ran a software company as CEO, stepping back from the code itself to focus on the business side. My fingers slowly forgot the muscle memory of brackets and semicolons as I traded development for spreadsheets, pitch decks, and strategy sessions.

But this experience with tools like Lovable, Manus, and Bolt has sparked something more profound than mere nostalgia. It's forced me to confront a question that every technologist, educator, and parent should be considering: In a world where AI can transform your verbal instructions into functioning software, is coding still a necessary skill?

Anthropic's CEO recently claimed that 99% of code will be AI-written within a year. Y Combinator reports that 25% of their startups are now generating nearly all their code using AI. The trend is undeniable, but the implications remain hotly contested.

The Sunset of Traditional Coding

The democratization of software creation represents nothing short of a revolution. Tools like the ones I've been experimenting with aren't just making coding easier – they're removing it entirely from the equation. This isn't cruise control for developers; it's a fundamental shift in who can build digital products.

The economic argument is compelling: businesses simply cannot justify spending weeks on what AI can produce in hours. When every startup is in a race against runway, this efficiency isn't just nice-to-have – it's existential.

This shift creates a paradox for junior developers. The traditional career ladder for software engineers typically begins with those simple, repetitive tasks that build muscle memory and internalize patterns. But these are precisely the tasks that AI excels at automating. How do you become a senior developer when the junior positions have evaporated?

The parallels to other technological transitions are striking. How many executives today understand the physical architecture of their computers? How many writers need to comprehend the mechanics of digital publishing platforms? Technology has always abstracted complexity, and code may simply be the next layer to disappear beneath the surface.

The Enduring Value of Computational Thinking

But there's another perspective worth considering, one that transcends the specific mechanics of writing code.

When I reflect on my software engineering classmates who have thrived over these two decades, their success wasn't built on remembering syntax or specific frameworks. What set them apart was a particular way of thinking – an ability to decompose complex problems, recognize patterns, and build systematic solutions. This engineering mindset travels well beyond the boundaries of software development.

Those with coding experience who embrace AI aren't merely becoming more efficient – they're experiencing a multiplicative effect on their capabilities. The combination of human judgment and AI implementation creates something greater than either could achieve alone. One developer I spoke with described it as "having an entire team of junior developers at my fingertips, but with the ability to direct them with precision."

There's also a profound difference between using an AI tool and being used by one. Those who understand what's happening behind the abstractions maintain agency in the relationship, recognizing both the capabilities and limitations of their digital collaborators.

The New Literacy Spectrum

Perhaps the most useful framing isn't whether coding is necessary or obsolete, but rather understanding where on the spectrum of technical literacy different roles and individuals might land.

I'm reminded of my university decision between computer science and software engineering. At the time, they seemed nearly identical, but twenty years later, the divergent career paths of my classmates reveal a crucial distinction. Those who emphasized the theoretical foundations found themselves equipped for research and science, while those who focused on engineering principles adapted into roles across the business spectrum.

The parallels to my experience with AI tools are striking. When I use Lovable or Manus, I'm not writing code, but I am engaging in a form of engineering – breaking down the problem, communicating requirements clearly, testing solutions, and iterating based on results. I'm in the backseat, but I'm still navigating.

This suggests a future where the value isn't in writing code but in thinking like someone who could. The abstraction level has shifted, but the fundamental need to understand how systems work remains.

The Great Reconfiguration

What we're witnessing isn't the end of coding but its transformation – from universal requirement to specialized skill, from mechanical execution to conceptual framework.

For parents wondering if their children should learn to code, for professionals considering whether to invest in technical skills, and for companies building their talent strategy, the answer likely lies not in abandoning computational thinking but in reframing how we develop and apply it.

The most adaptable among us will recognize that the real value has never been in the code itself, but in the mental models that coding helps us build. As AI abstracts away more of the implementation details, those mental models become more valuable, not less.

So perhaps the most important skill isn't coding at all, but the ability to understand which problems technology can solve, how to communicate our intentions to increasingly intelligent systems, and when human judgment must still prevail.

The question isn't whether coding is dead, but rather what new forms of literacy are being born in its place.

Unprompted is an occasional opinion column from Kunal Gupta for Pivot 5 readers. Follow Kunal on LinkedIn.