Taking a look into agentic AI and the emerging landscape around it

Web development is a fast-paced and ever-evolving field, with new technologies and frameworks emerging all the time. Since the arrival of AI tools such as ChatGPT, Claude, and GitHub Copilot, we've seen a sudden shift towards AI-driven web development, with these tools aiming to revolutionize the way we build and deploy web applications.

Naturally, questions have arisen in the tech community concerning the quality of agentic generated code, the ethics of using AI in development, and the impact it will have on the job market for web developers.

In this article, we'll explore the current state of AI-driven web development, the tools and technologies that are shaping the landscape, and what the future could hold for this field.

The current state of web development – moving to the cloud

According to W3Techs, platforms like Vercel and Netlify have completely changed how we build and host apps, making scalable cloud deployment almost effortless. Vercel alone now powers nearly 1% of all websites, almost double Netlify’s share.

With Next.js as its flagship framework, Vercel’s mission has been to free developers from tricky deployment scenarios and complex hosting issues. By offering a generous free tier, they’ve attracted everyone from hobbyists tinkering with side projects to start-up founders scaling their MVPs.

The old start-up adage, "Launch fast. Fix later.", has fueled this shift toward cloud-centric development. By removing many of the infrastructure headaches, developers are encouraged to focus their energy on shipping impactful features, rather than managing servers.

The reason to launch fast is not so much that it's critical to get your product to market early, but that you haven't really started working on it till you've launched. — Paul Graham, Startups in 13 Sentences

This has shifted the way we think about web development — prioritizing speed, iteration, and efficiency over managing infrastructure — which sets the stage for the rise of AI-driven web development.

Emergence of AI-driven web development - agentic / vibe coding

The emergence of agentic AI in web development — often referred to as vibe coding (for better or worse) — is reshaping how some developers approach building web applications.

Rather than manually writing every line of code, developers can instead describe what they want in natural language, with AI tools generating the scaffolding, logic, or even full components to bring that vision to life.

Here are some of the most popular tools in the space right now:

• ChatGPT: OpenAI’s conversational AI can generate human-like text and code, making it a versatile assistant for everything from scaffolding components to debugging and documentation.
• GitHub Copilot: GitHub’s AI-powered completion tool helps you write and refactor code faster, suggesting context-aware lines or entire functions to keep your workflow smooth and efficient.
• Claude: Built by Anthropic, Claude offers agentic capabilities for more complex workflows. It's particularly effective for multi-step coding tasks, refactoring, and planning projects when guided by precise prompts.
• Claude Code: A terminal-based AI agent that can analyze your project, generate TODO lists, and even commit code changes directly to your repository. It’s designed to help bring projects from local development to production with minimal friction.

AI-driven web development is still in its early stages, but what new AI users find most remarkable is how quickly ideas become reality — often in a fraction of the time compared to traditional methods. But with this speed comes new challenges.

The future of web development - proceeding with caution

While some in the tech community embrace this shift as a productivity revolution, others remain skeptical. According to a Stack Overflow Developer Survey 2024, just 43% of developers trust the accuracy of AI-generated code, while 31% remain skeptical, indicating a significant portion of the community doubts the quality and reliability of autogenerated outputs.

Completely trusting the output of AI for code generation comes with its own set of risks. Over-reliance on AI has been shown to lead to poorly optimized, insecure, or misunderstood codebases, making long-term maintenance difficult.

A recent TechRadar Pro report highlighted an “AI speed trap,” where the rush to release AI-generated code resulted in technical debt, service outages, and even million-dollar losses for some organizations.

As we are aware, when a project code grows in size, so does the complexity. This is often the point where developers find themselves running into issues, and where problem solving and analysis is needed to resolve the issues. An over-reliance on AI can lead to a lack of understanding of the underlying code, making it difficult to troubleshoot and fix issues when they arise.

What's important to remember is that AI tools are just that — tools. They can augment a developer's capabilities, but they can't replace the critical thinking, problem-solving skills, and domain expertise that human developers bring to the table. So how are developers adapting to this new landscape?

Redefining the role of the modern web developer

The shift could see a redefining of the modern web developer's role: instead of just being coders, future developers could be evolving into more of a role adjacent to designers, architects, and planners, capable of critically evaluating AI-generated solutions and ensuring that the end product is reliable, scalable, and secure.

Companies such as Microsoft, Google, and Meta are betting big on this new paradigm - with performance reviews now including criteria around AI tool usage and integration - using AI is no longer optional.

GitHub's CEO Thomas Dohmke recently stating in August 2025:

"`The software developer role is set on a path of significant change. Not everyone will want to make the change. Managing agents to achieve outcomes may sound unfulfilling to many, although we argue that’s what developers have been doing on a lower level of abstraction, managing their computers via programming languages to achieve outcomes. Still, as humans we are often reluctant to change, and that’s okay."
— Thomas Dohmke, CEO of GitHub Developers, Reinvented

Dohmke's perspective is interesting to consider that the developers he considers who will thrive in the new era are those who combine technical expertise, a holistic understanding of software and systems, and the flexibility to think strategically

The aim of thee tools is not to replace developers — far from it. The new level of abstraction that tech leaders are aiming for developers to adopt come in the form of managing AI agents to achieve outcomes.

How AI agents are reshaping development

To support this evolution, the industry has seen the rise of AI agents — tools that go beyond simple code generation to actively collaborate with developers throughout the build process. These agents help developers focus on higher-level strategic thinking rather than repetitive implementation tasks.

What makes AI agents unique is that they can:

• Analyze your project structure and suggest improvements.
• Automate repetitive setup or integration tasks.
• Coordinate between APIs, frameworks, and deployment pipelines for seamless workflows.
• Learn from your patterns to become more context-aware over time.
• Multiple agents can work together, each specializing in different aspects of the development process.

By offloading repetitive or low-level tasks to agents, developers are free to focus on higher-level thinking — such as system design, scalability, and ensuring that the product aligns with the bigger picture. This is the kind of shift that could define what it means to be a developer in the AI era.

AWS CEO Matt Garman is taking a pragmatic view of this evolution, and provides a useful framework for thinking about how developers can adapt:

"I think people would love to spend more of their time on that creative aspect, on the driving aspect of doing interesting analysis, on doing the parts of their job that they love and not the parts that they don't."
— Matt Garman, CEO of AWS

Given this rapid context shift, we should consider what are the tasks that we want to offload to AI agents, and what will that allow us to focus on instead?

Getting started with AI-driven web development

Exploring agentic coding doesn’t mean abandoning traditional development skills; it means augmenting them. As always, we encourage you to apply these situations to your own context and workflow, and figure out what works best for you.

Start small — try integrating these tools into your daily workflow:

• Use Copilot to refactor existing code or generate boilerplate - perhaps a blog that you never got around to finishing?
• Experiment with ChatGPT for generating components, testing scripts, or API integrations - perhaps create a design spec for a project in your head?
• Use Claude to generate documentation for your project - often times documentation is the last thing we do, and often the first thing to be neglected - Claude is really good at generating documentation from codebases.

To dive deeper, check out our project on Claude Code, where we walk through using Claude Code to create specific agents to help bring a local Next.js markdown website to production.

It's a hands-on way to see how agentic AI can supercharge your workflow while still keeping you in the driver’s seat.