Is AI The Death Of Coding?

In partnership with

Welcome to this week’s Deep-Fried Dive with Fry Guy! In these long-form articles, Fry Guy conducts in-depth analyses of cutting-edge artificial intelligence (AI) developments and developers. Today, Fry Guy dives into AI’s impact on coding. We hope you enjoy!

*Notice: We do not receive any monetary compensation from the people and projects we feature in our Sunday Deep-Fried Dives with Fry Guy. We explore these projects and developers solely to showcase interesting and cutting-edge AI developments and uses.*

Unlock the Social Media Tactics That Work Right Now

Is your social strategy ready for what's next in 2025?

HubSpot Media's latest Social Playbook reveals what's actually working for over 1,000 global marketing leaders across TikTok, Instagram, LinkedIn, Pinterest, Facebook, and YouTube.

Inside this comprehensive report, you’ll discover:

• Which platforms are delivering the highest ROI in 2025

• Content formats driving the most engagement across industries

• How AI is transforming social content creation and analytics

• Tactical recommendations you can implement immediately

Unlock the playbook—free when you subscribe to the Masters in Marketing newsletter.

Get cutting-edge insights, twice a week, from the marketing leaders shaping the future.

Download The Report

AI may be putting an end to the very practice that birthed it: coding.

Major tech companies are increasingly letting AI handle the heavy lifting in software development. In fact, Microsoft’s CEO Satya Nadella recently revealed that 20–30% of Microsoft’s code is now written by AI. On a recent earnings call, Google’s CEO Sundar Pichai similarly noted that over 30% of new code at Google is being automatically generated. These figures would have sounded like science fiction just a few years ago, but now they hint at a future where human coding might become the exception, not the norm.

It’s not just a few scripts or trivial code, either. Microsoft’s CTO Kevin Scott predicts that by 2030, “95% of all code” will be AI-generated. Some industry watchers are even more bold: Stability AI’s CEO Emad Mostaque argues that “there will be no programmers in five years,” citing GitHub data that roughly 41% of all code being written today is already written by AI. While those claims are provocative (and not everyone agrees with the exact numbers), there’s no doubt that AI coding tools have suddenly leapt from novelty to mainstream. In this article, we’ll explore why this is happening, what an AI-dominated coding future might look like, and why human coding skills may still matter in surprising ways.

THE END OF CODING?

Not long ago, conventional tech wisdom held that everyone should learn to code. Today, some tech leaders are reversing that advice. Nvidia’s CEO Jensen Huang declared that with modern AI, “Everybody in the world is now a programmer. This is the miracle of artificial intelligence.” Huang argued that, thanks to generative AI’s rapid advancement, learning syntax like Java or C++ (fancy coding languages) is no longer a priority. Essentially, anyone can be a programmer by describing what they want in plain language—AI will take it from there.

Since AI will handle all the coding, Huang suggested that time spent learning how to code would be better used gaining expertise in other domains. This startling viewpoint—essentially “don’t learn to code”—has been echoed by others. John Carmack, the famed engineer behind Doom, agreed that the real value lies in problem-solving, not typing syntax. “Coding was never the source of value ... Problem solving is the core skill,” Carmack noted, adding that the discipline learned in programming remains useful even if AI winds up better than human programmers. In other words, thinking like a programmer may be important, but writing actual code could become a niche activity.

Of course, not everyone is ready to throw in the towel on human coding, and many experts think it will remain highly relevant in the future. But the fact that industry heavyweights are seriously debating whether people should keep learning code at all shows just how profoundly AI is shaking up the field.

WHY IS AI CODE BOOMING?

Why is this shift happening now? The short answer: today’s AI coding assistants have become astonishingly good and widely available. Tools like OpenAI’s Codex, GitHub Copilot, Amazon CodeWhisperer, and Google’s Gemini can generate substantial chunks of code from a simple prompt or even continue writing code from where you left off. They act as tireless assistants for developers, autocompleting functions or producing entire modules upon request.

At their core, these AI models are large language models (LLMs) trained on vast swaths of existing code. For example, OpenAI’s Codex was trained on billions of lines of public code from GitHub, learning the patterns of languages like Python, JavaScript, C++, and more. Essentially, if you give Codex (or a similar model) a description of what you want or the beginning of a function, it predicts code that likely achieves that—much like an advanced autocomplete, but powered by knowledge of practically every open-source project out there. GitHub Copilot, which integrates into code editors, uses this technology to suggest the next line or entire blocks of code as you type. Developers have found it uncannily good at boilerplate tasks. For instance, it can flesh out a routine to parse a file or create a web interface from comments, saving time and effort.

Like all AI models, these coding assistants don’t think in the way humans do, but they’ve seen so many examples that they can often produce correct (or at least plausible) solutions to common programming problems. And they’re improving fast.

The appeal for companies is obvious: productivity can skyrocket. GitHub’s own research found that developers using Copilot write code 55% faster on average. Routine chores like writing unit tests or translating one programming language to another can be offloaded to the AI, freeing human developers for higher-level design work. In an era of intense demand for software, automating chunks of coding can save money and time. Even if the AI’s output isn’t perfect, it gives a huge head start to developers. It’s an assistant who never sleeps, offering suggestions for every task.

There’s also a broader democratization at play. With AI helpers, people with little coding experience can create software by describing what they need in natural language. This lowers the barrier to entry—a scientist or artist can prototype an app without having to deeply learn programming. Tech visionaries see this as opening the floodgates of innovation, as more people can turn ideas into working software through AI intermediaries.

A FUTURE WITH (ALMOST) NO HUMAN CODERS

What would a world look like where developers write little to no code by hand? It’s a scenario that seemed far-fetched until recently. Now it’s plausible to imagine a software project where a human’s job is to explain the desired outcome to an AI, and the AI model writes all the code—from backend logic to front-end design. In this future, a single engineer might oversee a team of AI “coders,” guiding them with prompts and high-level instructions rather than writing code themselves. Software development could become more about curating requirements and testing outputs than about typing out algorithms.

New programming paradigms might emerge. Instead of writing Java or Python, future developers might craft intricate prompts in English—essentially “programming” the AI by describing features and constraints. The AI could iteratively refine the product, perhaps even debugging and optimizing itself through additional AI feedback loops. The role of the human might resemble a product manager or editor, setting goals and verifying results, rather than a craftsperson writing each line.

Such a world might also blur the line between end-users and programmers. If anyone can create an app by telling an AI what they want, we could see an explosion of custom software built by non-engineers. Need a simple mobile game or a data visualization for work? Just ask your AI assistant, and it can generate it. This is the utopian vision behind the “no-code” movement, supercharged by generative AI. In practice, we’re not fully there yet—but we’re inching closer every day as AI models improve in reliability.

It’s important to note that even the most optimistic AI proponents acknowledge some caveats. Today’s AI coding assistants can make mistakes or produce insecure code if left unchecked. They sometimes need guidance to get complex logic exactly right. In the near-term future, humans would still be in the loop, monitoring and refining what the AI produces—much like an editor correcting a rough draft. But over time, as AI gets more sophisticated, it’s conceivable that the human oversight needed will diminish for many standard programming tasks.

A HUMAN-AI CODING PARTNERSHIP

In conclusion, the rise of AI in programming doesn’t spell doom for humans—it augurs a new partnership. Just as calculators didn’t eliminate the need to learn math, AI coders will change how we create software. Routine coding might be largely automated, but human creativity, oversight, and domain expertise will guide the AIs in building something meaningful and safe. The job title “coder” might fade, but “software architect,” “AI composer,” or “machine collaborator” could take its place.

Change is coming fast—faster than many of us imagined—but it’s an exciting time. We stand to gain incredible productivity and open the world of software development to more people than ever. Rather than eliminate the need for human programmers, AI can elevate developers to work at a higher level of abstraction. In this new era, knowing some coding is still useful—not because you’ll spend all day coding, but because it empowers you to better direct and utilize our new AI coding partners. The tools might be different, but the ultimate goal remains: solving problems and building great things. And humans are very much still in that picture, even if we’re no longer writing every line ourselves.