Introduction

Product design is at a crossroads. After attending PUSH UX in Munich—one of Europe’s largest gatherings for user experience and product design professionals—I noticed a clear pattern emerging among attendees from companies like Anthropic, Wise, Google, and IKEA. Teams are no longer asking whether to use artificial intelligence. Instead, they are dividing into two distinct groups: those who treat AI as a cautious experiment layered onto existing processes, and those who have woven it directly into how they research, prototype, design, and ship. This split is about more than which tools appear in a designer’s toolbar. It reflects deeper questions about how product teams collaborate, where value is created, and who owns the final outcome.

What Happened

The conference surfaced a tension that many product organizations are feeling right now. Paulo Oliveira, Global Innovation and Technology Manager at IKEA, warned against the instinct to lead with AI tools and automation. He argued that faster cycles do not automatically equal value, and that piling up unused pilots can leave teams wondering where their core strengths went. His point was that meaningful adoption starts with identifying where time is actually lost, which handoffs create friction, and how AI can improve research, design, building, and validation—not just speed up individual tasks.

At the same time, the role of the product designer itself is shifting. For years, design and engineering lived in separate worlds: designers worked in Figma, wrote specifications, and handed everything off to developers. AI-assisted development is narrowing that gap. I spent the past year moving from almost entirely Figma work to building applications directly with AI coding tools. This change let me design, prototype, and implement interfaces in a continuous loop, reducing the number of handoffs between design and engineering. For product teams, that opens the door to connected workflows, stronger designer ownership, leaner teams, and the ability to stretch into areas like search optimization and app store visibility. This does not mean every designer must become a software engineer; deep specialization still matters, especially for complex products. But working closer to implementation allows designers to contribute beyond traditional role boundaries.

Prototyping is another area where AI is changing the game. Creating interactive mockups for specific hardware, such as an Apple Watch or an in-car infotainment system, used to require significant engineering support. Now designers can generate device-specific interactive prototypes without writing production code, making it easier to test ideas in realistic contexts.

Why This Matters

As AI makes exploration and execution faster, the focus of design work is moving toward judgment and quality. Knowing whether a solution actually solves the right problem, serves real user needs, and performs reliably still requires human experience and critical thinking. Julia Bastian, Product AI Lead at Alasco, addressed this directly in a session on low-quality AI-generated research—sometimes called discovery slop. She emphasized that individual productivity gains are not enough. Organizations need shared workflows, clear quality standards, and a disciplined way to evaluate AI-generated output.

There is also a cultural dimension. Building an AI-first product culture means encouraging experimentation without sacrificing technical rigor, security, or reliability. This is especially important for teams working on deep technology, where mistakes can be costly. The quality of AI-assisted results depends on how thoughtfully people guide, review, and refine the process. Sometimes AI-generated experiences fall short, but that gap can close with better oversight and stronger design instincts.

Key Takeaways

  • Product design teams are splitting into two camps: cautious experimenters and teams that integrate AI across the full workflow.
  • Adopting AI tools does not automatically create value; faster cycles can mask unchanged processes and unused pilots.
  • AI-assisted development lets designers work closer to code, reducing handoffs and increasing ownership of the final product.
  • Designers do not all need to become engineers, but they can contribute beyond Figma by participating in implementation and validation.
  • Interactive prototyping for specific devices is now more accessible without heavy engineering support.
  • Product quality and decision-making matter more than raw speed, requiring clear standards and human judgment.

Conclusion

The design community is still figuring out where AI fits. Some teams are finding their footing, while others have already restructured how they collaborate. My own experience has made me more comfortable moving between design, product strategy, and engineering, but there is no single correct path for every organization. The most important shift may be the willingness to reconsider established workflows, look for places where new capabilities genuinely improve outcomes, and take greater responsibility for business results. AI is not a replacement for design thinking; it is a tool that rewards teams willing to rethink how they work.