The rise of vibe coding

Inside the shift toward conversational, AI-assisted programming

Image credit: Sora and Canva

In early 2025, AI researcher Andrej Karpathy introduced a term that quickly gained traction in the software development community: vibe coding. Originally intended as a casual expression—referring to working fluidly with AI development tools—it has since come to describe a broader shift in how developers interact with code.

Rather than focusing on syntax and structure line by line, vibe coding centers on a collaborative, prompt-driven workflow where human developers guide AI systems using natural language. The result is a faster, more conversational style of development—one where the code often emerges through interaction rather than manual construction.

What is vibe coding?

Vibe coding generally refers to the practice of working with AI-based development tools in a more intuitive, high-level manner. Instead of writing every line of code manually, developers describe functionality or changes in plain language, and the AI generates code in response.

For example, rather than adjusting CSS styles directly, a developer might say:
“Reduce the sidebar padding by half,” and the AI interprets and implements that instruction.

This approach reflects a move toward prompt-oriented programming, where the emphasis is less on low-level implementation and more on collaborative problem-solving between humans and machines.

Image credit: Substack image generator

A brief history

While the term "vibe coding" is new, the trajectory toward AI-assisted development has been underway for several years:

• 2018: early AI tools like Tabnine introduced predictive code completion.

• 2021: GitHub Copilot (powered by OpenAI Codex) made AI-assisted programming widely accessible.

• 2022–2024: tools such as Amazon CodeWhisperer, ChatGPT, and Google's Gemini brought conversational capabilities to the developer’s toolkit.

• 2025: the term “vibe coding” becomes widely used, popularized through social media and developer forums.

These advancements have contributed to a shift in how many developers think about coding—not just as a mechanical task, but as a higher-level design process supported by AI.

Image credit: Substack image generator

Developer experience and productivity

Early feedback from the software development community has been largely positive, especially regarding productivity. Several studies and internal reports have found that developers using AI tools can:

• complete tasks more quickly (with productivity gains reported up to 55%).

• reduce time spent on repetitive or boilerplate code.

• maintain focus by offloading lower-level implementation details.

That said, effective use of these tools still requires a strong grasp of software architecture, debugging, and quality assurance. AI can accelerate coding—but not eliminate the need for thoughtful human oversight.

Tools enabling the shift

A growing number of tools support this prompt-based development approach:

• GitHub Copilot Chat

• Amazon Q

• Replit Ghostwriter

• Cursor IDE

• Klover AGD

• Lovable.dev

These platforms integrate AI directly into the developer workflow—either in the IDE, via chat interfaces, or through web-based low-code environments. Many of them aim to make software creation accessible not only to engineers, but also to designers, product managers, and technical generalists.

Image credit: Substack image generator

Challenges and limitations

While the potential of AI-assisted development is significant, there are limitations:

• Code quality: AI-generated code may function correctly but lack maintainability or clarity.

• Debugging complexity: developers may need to spend time understanding and verifying unfamiliar AI-generated logic.

• Security risks: automatically generated code can introduce vulnerabilities if not properly reviewed.

• Team integration: managing and reviewing AI-produced code can raise collaboration and documentation challenges.

Several organizations, including IBM, have expressed concern about overreliance on AI-generated code without sufficient human verification or architectural planning.

Looking ahead

Vibe coding reflects an ongoing evolution in software development, where natural language and AI tools are playing an increasingly central role in the creation process. Whether used for prototyping, refactoring, or full-stack development, these tools offer new possibilities for speeding up workflows and lowering technical barriers.

However, as with any emerging methodology, critical thinking, review processes, and best practices remain essential.

For now, vibe coding may not replace traditional development, but it is clearly shaping the future of how code is written—and who gets to write it.

Image credit: Substack image generator

📚 Bibliography

• Business Insider. (2025, June). Tech companies have a new hiring requirement: vibe coding. https://www.businessinsider.com/vibe-coding-tech-firms-hire-engineers-2025-6

• Business Insider. (2025, February). Silicon Valley's next act: bringing 'vibe coding' to the world. https://www.businessinsider.com/vibe-coding-ai-silicon-valley-andrej-karpathy-2025-2

• Dany Kitishian. (2025, May). Google Gemini: “vibe coding” uproar – navigating the realities of AI-assisted software development. Medium. https://medium.com/@danykitishian/google-gemini-vibe-coding-uproar-7fc4d95826c1

• IBM. (2024). Are new‑gen AI tools putting your business at risk? https://www.ibm.com/think/news/are-new-genai-tools-putting-your-business-at-risk

• Karpathy, A. [@karpathy]. (2025, February 2). There’s a new kind of coding I call “vibe coding”... [X post]. https://x.com/karpathy/status/1886192184808149383

• Peng, S., Kalliamvakou, E., et al. (2023, February 13). The impact of AI on developer productivity: Evidence from GitHub Copilot. arXiv.
  https://arxiv.org/abs/2302.06590

• Pragmatic Engineer (Orosz, G. & Nilsson, D.). (2025, June). Vibe coding as a software engineer. https://newsletter.pragmaticengineer.com/p/vibe-coding-as-a-software-engineer

• Tanium. (2025, May 20). What is vibe coding? The pros, cons, and controversies.
  https://www.tanium.com/blog/what-is-vibe-coding

• Visual Studio Magazine. (2024, September). Another report weighs in on GitHub Copilot dev productivity. https://visualstudiomagazine.com/articles/2024/09/17/another-report-weighs-in-on-github-copilot-dev-productivity.aspx

• Wired. (2025, June). Vibe coding is coming for engineering jobs. https://www.wired.com/story/vibe-coding-engineering-apocalypse

More news!

OpenAI secures $200 million Pentagon contract to develop military AI