“Vibe coding,” will it vibe forever?

As the world advances in AI with each passing moment, we encounter new terminology all the time—one of the latest being “vibe coding.” But is it all about the vibe, or does actual coding take place? With new AI models being launched almost daily, the tech ecosystem is bustling with innovative concepts and trendy methods of working.

In the recent times the cyberspace is bustling with this new term, coined by Open-AI co-founder Andrej Karpathy and he said:

AI tools have made everyone’s life easier, so why not people who code, or maybe willing to code without having a technical background. Therefore gone are the days, when someone has to have a technical qualification to code.

The Rise of AI-Driven Coding

It is an emerging and trendy approach to software development that emphasizes an informal, conversational, and intuitive method of coding.

Amidst the rise of Large language models (LLMs) , vibe coding has become smoothly accessible to even non-technicals using tools like Cursor AI and Windsurf AI, where the user acts as a guide, directing the AI agents on what to develop.

The essence of this new method is to interact with the AI agents via voice prompts and instructions rather than directly manipulating the coding ecosystem.

Even though companies like Accenture, IBM, Replit have been using vibe coding for years, assisting young coders and early professionals with autocoding, some startups have brought in newer approaches and tools to facilitate autocompletion of coding not just for professionals but also amateurs.

Codeium’s Windsurf and Anysphere’s Cursor

Two of the trendsetters in this arena are Codeium’s Windsurfer and Anysphere’s Cursor, pushing the boundaries, integrating AI-driven capabilities into coding environments.

These tools are becoming sophisticated agents to develop softwares with minimal or no direct code interactions.

In simpler words of a user, “ you tell AI exactly what you wanna build and it builds it for you.” 

While remarking on how he uses vibe coding, the user also mentioned the voice-to-text application that comes into play while interacting with the AI-powered agent. 

Instead of laboring over every line of code with meticulous writing, reviewing, and debugging, developers harness the power of AI tools to effortlessly handle much of the heavy lifting.

Codeium via its Windsurfer editor incorporates AI to enhance coding activity by providing intelligent suggestions, refactoring and debugging.

It offers access to high-level AI models that acts like a copilot and handles complex tasks independently.

According to the team, the Windsurf Editor is where the work of developers and AI truly flow together, allowing for a coding experience that feels like literal magic.

But what made the formation of Windsurf possible?

The development of an AI-powered code editor began with a fundamental question: What features should such an editor include, and what should it avoid, if it is capable of searching, traversing, and analyzing code, as well as making edits across multiple files and executing terminal commands?

They quickly realized that they needed to create something different from what all standard code assistants offer, such as chat-based functionality. This approach could completely transform the editor’s interface from a model-driven focus to a user-centered experience. 

Thus, the Codeium Chat Cascade was developed. This system is designed to iterate on itself and can answer broader questions or tackle larger, more complex tasks compared to an AI limited by a single LLM inference.

A notable feature of Windsurf is its clear distinction between directly interacting with the codebase—such as modifying a specific file like app.js—and engaging in general conversations with the AI. When a file is selected in Windsurf, the generated code is tailored specifically for that file, while the chat function is designed for general questions and offers assistance in understanding various concepts. 

The team of Windsurf was long aware that the AI code assistant space has been around long enough and is now swamped. If a tool doesn’t enhance developer productivity, it won’t be part of the discussion. The main objective is to enable anyone to accomplish tasks they couldn’t do before and that is where the concept of “vibe coding” has vibed with both software professionals and amateurs looking to develop something.

This success has achieved them a unicorn status after it raised $150 million in Series C funding, increasing the company’s valuation to $1.25 billion in less than two years of its inception. The funding round was led by General Catalyst, with ongoing support from existing investors Kleiner Perkins and Greenoaks.

What is it that Windsurf is doing differently than its competitors?

While both Windsurf as well as its competitor Cursor offer real value to developers in today’s software ecosystem, their features, accessibility and affordability says otherwise. 

Even though their functionality  are quite similar, but their affordability is not quite so much, Cursor AI’s first paid tier is around $20 USD, while Windsurf’s is $10 USD . The free tier of Cursor AI has reduced features due to increased user demand, causing credits to deplete quickly. In contrast, Windsurf Editor offers higher quality code outputs for its free tier.

Coming to the models used, both the IDEs offer access to high-level AI models like GPT-4, Claude 3.5 Sonnet, but some additional models in Cursor AI’s free tier do not provide outputs as good as Cascade-Base, which is accessible in Windsurf’s free tier.

But vibe coding has not remained restricted with these two AI-agents, Venture capitalists and developers have been looking forward to some other AI-powered coding tools from Bolt.new, Lovable, Pica. 

Will It Replace Software Engineers?

The hype surrounding vibe coding continues to grow, but there are important questions about its long-term impact. Will vibe coding ultimately replace software engineers? According to a recent study, software engineers who utilize code-generating AI systems are more likely to introduce security vulnerabilities into the applications they create.

Even though it is a much-hyped technique for software developers to carve a path in this tech ecosystem, some are concerned whether it will voraciously consume the jobs of data and software engineers. While some believe it works for simpler projects, others argue that for complex developing applications, it may still require greater training and feedback from human developers.

These are all trial and error methods unless some substantial experiment takes place, which could once and for all answer the question: Will vibe coding replace traditional coding or will it become just another tool to assist developers in creating software more efficiently?

Vibe coding represents a radical shift in how software development is being approached, driven largely by the power of AI. It democratizes coding, making it accessible even to those without technical expertise. But whether it will eventually replace the role of the traditional software engineer is still an open question. One thing is certain, though: AI-powered coding tools like Windsurf and Cursor are shaping the future of software development, making coding more intuitive, efficient, and accessible to a broader range of users.