
Future Frontend: Rethinking How We Build for the Web
Frontend development is evolving quickly. New frameworks, browser capabilities and AI-powered tools constantly reshape how digital products are built and experienced.
To explore some of these changes firsthand, Ruben Diogo, Frontend Engineer at Mercedes-Benz.io, attended Future Frontend, a conference bringing together developers, designers and technology experts to discuss where the web is heading.
While the event covered a wide range of topics, one idea kept surfacing throughout the talks and discussions: as AI becomes increasingly integrated into digital experiences, developers may need to rethink some of the assumptions that have shaped web development for years.

Ruben Diogo
Contents
Looking Beyond the Browser
The talk that had the biggest impact on Ruben came at the very end of the conference.
Rachel Lee Nabors introduced the concept of the Headless Web, with the idea that websites are no longer consumed exclusively through visual browsers. Increasingly, content is accessed by APIs, automations and AI-powered systems that interact with information without ever rendering a page the way humans do.
For Ruben, this perspective changed how he thinks about frontend development.
I went to Future Frontend trying to understand where frontend development is heading. This was one of the talks that genuinely changed how I think about the web.
The discussion raised an important question:
Are we building surfaces that only make sense to humans? And if so, what does it cost us as agents become a more common entry point into digital products?
As AI systems become increasingly capable of understanding and interacting with digital products, things like semantic HTML, accessibility standards and structured content may play an even more important role and not only for people, but for machines.
Could Web Architecture Be Due for a Shift?
Another session that challenged Ruben’s thinking focused on Hypermedia-Driven Architectures (HDAs).
The presentation argued that modern web applications have gradually pushed increasing amounts of complexity into browsers, while alternative approaches could make interactions more discoverable and explicit.
What interested Ruben most wasn’t necessarily the architecture itself, but the broader implications.
What stood out to me was the idea that some of today’s frontend decisions might actually make it harder for non-human clients to understand and operate systems later on.
At the same time, he believes the approach still raises valid concerns.
Most frontend engineers are already familiar with frameworks such as React, Vue and Angular, so adopting a completely different approach wouldn’t be straightforward.
For Ruben, the session wasn’t about finding the next dominant architecture. It was about questioning whether the assumptions behind today’s architectures will still hold true in a world where AI agents become increasingly common users of digital products.
Small UX Decisions, Big Impact
Not every takeaway revolved around AI.
A presentation by Una Kravets focused on motion, UI patterns and emerging browser capabilities.
One of the key messages was that animations shouldn’t be decorative, they should help users understand what’s happening, guide attention and make interactions feel more natural.
For Ruben, the practical nature of the session stood out immediately.
During the presentation, I found myself applying some of these principles to experiences we already work on.
The talk also explored technologies such as Anchor Positioning and reinforced broader design principles around reducing noise, respecting user preferences and creating more intuitive experiences.
Building Software That Can Adapt
Another memorable presentation came from Matthew Mamonov, who explored the relationship between refactoring and architecture.
His central argument was that cleaner code doesn’t automatically mean better architecture. Instead, architectural health often reveals itself through qualities such as testability, maintainability and adaptability
One idea in particular stayed with Ruben.
If your tests are awkward, it might not be a testing problem. It could be a design problem revealing itself.
The session served as a reminder that great software isn’t just about solving today’s challenges. It’s about creating systems that can evolve as requirements change.
Why Intent Matters More Than Ever
AI also featured heavily in discussions around software development itself.
During a session on agentic development, Ohans Emmanuel introduced a principle that resonated strongly with Ruben: intent is the first-class artifact.
The idea is simple but powerful. If requirements are vague, AI systems fill the gaps with assumptions. Clear intent, constraints and expected outcomes create better results.
The principle that stayed with me was that intent needs to be defined with the same rigour as acceptance criteria. Otherwise, AI fills the gaps with assumptions.
Later, during a workshop on AI agents, Ruben explored how autonomous systems can perceive context, make decisions and execute actions to achieve specific goals. The session offered a practical look at how these technologies could influence future development workflows.
One Takeaway That Connected Everything
Although the conference covered a broad range of topics, from UX and architecture to AI and emerging web standards, Ruben left with one recurring thought.
The conversations weren’t really about frameworks, animations or AI tools.
They were about how the web itself may be changing.
As we enter a new era of web development, where many things remain uncertain, one thing is clear: agents are here to stay. The future of building for the web is about creating systems that humans and agents can understand, interact with and navigate seamlessly.
Whether discussing architecture, accessibility, user experience or development practices, many of the conference sessions pointed towards the same challenge: building digital products that work well for the users of today while preparing for the users of tomorrow.
And according to Ruben, that conversation is only just beginning.
Related articles

Mariana Pereira
Learning to Let Go: What Leadership Ateliers Lisbon Taught Me About Influence, Impact and Growth
What does leadership look like when your role isn't defined by managing people, but by bringing different perspectives together?
Oct 2, 2026

Vaibhav Khatri, Nikolaus Kober
5 Skills Every AI Engineer Needs Today (And Why the Role Matters More Than Ever)
From system design and evaluation to Responsible AI and observability, here's what it takes to build AI that works in the real world.
Sep 29, 2026

Ghazaleh Motia
What n8n Taught Me About Product Prototyping
Product ideas often can be expensive to validate. Before something becomes a real product, teams/individuals frequently need to answer a simple question: does this idea actually solve a problem worth tackling? Over the l
Sep 18, 2026