Web Directions Year round learning for product, design and engineering professionals

The conversation around AI and web development has moved far beyond code completion and chatbots. At Web Directions Developer Summit, we’re exploring something more fundamental: how AI is reshaping not just our tools, but our entire conception of what we’re building and who we’re building for.

Building Agents: Less Magic, More Engineering

The mystique around AI agents is both a barrier and an opportunity. Kevin Yank from Culture Amp cuts through the illusion in his session “How to Build an Agent in JavaScript.” Based on Thorsten Ball’s original work, Kevin demonstrates that a functional coding agent—the kind powering tools like Cursor, Windsurf, and Claude Code—requires roughly 400 lines of JavaScript. No magic, just engineering.

This matters because understanding how agents work under the hood changes how we reason about when and how to use them. When you’ve built one yourself, the limitations become as clear as the possibilities. You stop treating these tools as black boxes and start making informed architectural decisions.

Making AI Actionable

AI that merely responds is already outdated. Tamas Piros explores this transition in “When AI Stops Talking and Starts Doing.” Through the Model Context Protocol, AI systems can now execute file operations, manage digital assets, analyse images, and automate workflows that previously required human intervention.

This isn’t theoretical. Tamas will demonstrate how to connect AI models to existing developer tools and services safely, combining structured function calls with natural language interaction. The result is AI that doesn’t just advise—it acts. This fundamentally changes the role AI plays in development: from passive assistant to active collaborator.

The Protocol That Changes Everything

Anna McPhee from Automattic has been deep in the trenches of Model Context Protocol integration. Her session “Teaching Your Frontend Tools to Talk to AI” examines what’s actually working in production—React component generation, Figma-to-code pipelines, and the growing ecosystem of MCP servers that frontend developers can use today.

More importantly, Anna shares what she’s learned about building frontend-focused MCP integrations that developers actually want to use. The patterns emerging across successful implementations, the pitfalls to avoid, and the opportunities still unexplored. This is practical knowledge from someone shipping these integrations at scale.

Beyond the Browser

Perhaps most provocatively, Rupert Manfredi argues we’re witnessing a platform shift that could displace the web itself. His session “What’s Beyond the Browser: The AI Platform Shift” examines how AI clients will fundamentally change what we mean by an “app.”

When AI can interact directly with services, when natural language becomes a primary interface, when agents handle complex workflows autonomously, the traditional browser-based web becomes just one possible interface among many. Rupert’s Unternet project isn’t science fiction—it’s building the open protocols for this emerging ecosystem. The question isn’t whether this shift is coming, but whether it develops in the open or behind proprietary walls.

Machines as Customers

Katja Forbes asks a question that forces us to confront the full implications of this transformation: what happens when your users aren’t human?

In “Machines Are Customers Too,” Katja explores how AI agents are becoming independent economic actors, making purchasing decisions without human intervention. Traditional customer journeys, trust-building strategies, and engagement patterns all require fundamental rethinking when the customer is code and loyalty is algorithmic.

This isn’t distant speculation. The first wave of AI agents making autonomous purchasing decisions is already here. The businesses that adapt their customer experience for both human and machine customers will have years of advantage over those that don’t.

The Design System Connection

The AI revolution also intersects with design systems in unexpected ways. Tammie Lister explores this in “From Components to Prompts,” examining how AI has accelerated the need for open, systematic design approaches. Mandy Michael demonstrates how tools like Figma MCP Server are already bridging the designer-engineer gap through AI-mediated workflows.

Why This Matters Now

A year from now, understanding how to build with AI, for AI, and alongside AI will be table stakes for professional developers. The teams that master Model Context Protocol, that understand how to architect for both human and machine users, that can build and evaluate AI agents—these teams will lead.

But right now, while these patterns are still emerging, you have the opportunity to learn from practitioners who are already shipping these technologies. To understand the architecture, the protocols, and the paradigm shifts while they’re still taking shape.

This is your chance to be ahead of the curve, not scrambling to catch up.

Register for Web Directions Developer Summit and join us November 19-20 in Sydney (and streaming) to explore the AI revolution with the people building it.