You’ve been described as the “world’s most high-resolution architect.” How do you use technology in your practice to develop and deliver radical new approaches to architecture and urban design?

For over two decades, I’ve been working in architecture, focusing on computational and algorithmic multi-agent systems, as well as artificial intelligence. My primary interest lies in engaging with complex systems—what I refer to as ‘architecture for complexity’. The central question I’ve been exploring is how to address the challenges within built environments, which I see as dynamic, interconnected ecologies, through the lens of computation as a connective tool.

What excites me most is the potential of computation and new technologies that enable us to engage with architecture, cities, urbanism, and infrastructure in ways that were once unimaginable. For the first time in history, we have access to big data, higher resolution, and computational power that allows us to reveal previously invisible patterns and relationships. In science, increased resolution makes the invisible visible—and this concept has been central to my architectural practice. The connective power of computation enables the synthesis of systems, revealing new insights and opportunities in design.

You once called cities the most complex systems humans have ever created. How can we use algorithmic design tools to synthesize this complexity and inform the design of more intelligent urban plans?

Cities—and those designing them—must undergo a fundamental shift in mindset. Today, we can simulate weather conditions at a one-kilometer resolution, track mobility patterns with precision, and model robotic construction in unprecedented detail. The ability to test complex scenarios in controlled environments is transformative, yet architecture and urban design have yet to fully engage with the depth of these possibilities. Buildings and cities are intricate systems that demand synthesis across multiple logics and data streams—far beyond isolated algorithms or aesthetic intuition.

Algorithms cannot exist in a vacuum; they must interact with external forces, integrating environmental, contextual, and systemic factors rather than merely reflecting an architect’s vision. To navigate complexity, architects must adopt a mathematician’s mindset, operating within an n-dimensional design search space instead of conventional three-dimensional thinking. Machine learning enables us to explore this expanded design space, uncovering solutions once beyond our reach. Yet, the essential question remains: How do we, as humans, harness creativity to guide this search—collaborating with machines to reveal the underlying patterns that shape our cities?

How did you start to apply these concepts to three-dimensional built structures?

After experimenting with waves, multi-agent systems, swarming behavior, and robotics—testing construction principles through prototypes like chairs and tables—I realized architecture needed a fundamental rethink in response to the planetary crisis. To stay relevant, we as architects must engage with systemic challenges and hack complexity to address urgent global issues.

Moving beyond high-resolution objects like lattice structures, I explored how to create lightweight yet strong systems inspired by nature. This led to larger questions: How could landscape infrastructures and prefab architecture—long defined by repetition and monotony—be reimagined with greater richness?

The “AI Timber” project emerged from this inquiry, using repetitive elements to form adaptable structures. Computational principles reveal that even simple building blocks—like binary code—can generate immense complexity. These discrete elements can produce formulas, images, music, or architecture, forming systems that respond to external forces while maintaining internal logic. I focused on the joints—how machine learning and algorithms could extend human cognition, revealing new patterns and generating complexity from repetition. By adjusting a single angle, every component could be unique while using the same base elements. Higher resolution allowed parts to respond dynamically to light, temperature, and other factors, all within a stable underlying system. The result is a structure that appears to evolve organically while maintaining coherence—offering vast possibilities for future applications.

Many of your projects have a distinctive aesthetic expressed through fluid, nebulous forms based on complex adaptive systems found in nature. How can these ideas be scaled up for use in an urban or infrastructural context?

Nature’s complexity has always been a deep source of inspiration. Desert dunes and rock formations—like those in the Grand Canyon—manifest intricate patterns that feel like vast computations of physics and matter. This notion of embedded complexity shapes my design approach, much like composing music: first assembling instruments and then exploring their interplay. In computational design, I often introduce “noise” sparingly—subtle disruptions that generate patterns beyond simple scaling functions. The goal is to use computational tools to uncover hidden forces and dynamics within materials, allowing aesthetics to emerge organically from these underlying systems.

My project, “Filling the Void,” applied these ideas to solar fields, transforming abandoned mining sites in the Australian desert into beautiful and smart site-specific solutions. Too often, these large-scale infrastructure projects are devoid of design and become ugly pieces of machinery that spoil beautiful landscapes. Designers should always be part of the process—not just to make them aesthetically pleasing, but to synthesize all the different elements involved.

The tools that enable this collaboration must evolve, too. Architects, engineers, and builders still operate in fragmented digital ecosystems, relying on incompatible software. That needs to change. With agentic AI, we can synthesize ideas more fluidly. Platforms like Omniverse are already allowing real-time simulations of complex scenarios, accelerating the design process through generative AI. Multimodal AI that integrates language, visuals, video, and 3D further expands these capabilities. The real challenge is synthesis: how do we merge these different modes of intelligence and think creatively?

Architecture has always been about integration—balancing client needs, site conditions, budgets, materials, technology, and cultural context. Now, with computational power and AI, this ability is amplified. The future of design isn’t just about individual buildings but about circularity, life cycles, and ecology—engaging complexity to create more intelligent, adaptive environments.

Some architects and designers today are using tools like artificial intelligence to generate fantastical images of structures that defy physics and can, therefore, never be built. How important is it to focus instead on using these technologies in ways that will genuinely benefit society?

The true potential of tools like AI lies in directing that creativity toward the real complexities—into matter, into physics, and into the systems that govern how we build. The focus shouldn’t just be on generating images; it’s about pushing beyond the surface, beyond the lowest common denominator. After a while, AI-generated images feel repetitive, lacking the spark of true innovation. That comes when we push these tools further with skill and intent. We should absolutely embrace them, but we need to be more ambitious—using them not just for aesthetic exploration but to reshape how we design, build, and think about cities as interconnected, evolving ecologies.

Historically, the most impactful architecture—whether Gothic or Modernist—was deeply connected to the context and the moment in which it was created. Today, we find ourselves at a similar crossroads, with the pressure to think boldly more urgent than ever. Rather than focussing on objects, form, and beauty, as many architects today continue to do, we need to consider how these elements link intrinsically with the world we live in and the challenges we must confront. Of course, aesthetics is essential but it’s tied to something deeper. I’ve always been interested in capturing what was once invisible—something computationally and physically complex but unseen. The aesthetic is a byproduct of that deeper exploration, not the end goal. We also need to move away from predominantly designing singular buildings to designing systemic frameworks—rules and protocols that allow architecture to emerge dynamically rather than being predetermined. The future architect needs to be more of a systems designer than a form-maker.

One risk I see with AI-driven design is flattening. After encountering enough AI-generated images, they all begin to blur together, losing their uniqueness. While AI can undoubtedly inspire creativity, there’s a risk of it becoming formulaic and detached from the real world. Thinkers like Nick Bostrom have raised important concerns about optimization. If AI is trained to focus solely on "optimization," it could prioritize efficiency over human sensibility and creative intent. For designers, the real opportunity lies in harnessing our creative power alongside AI, ensuring we retain that irreplaceable human touch—the emotional intelligence that gives our designs true meaning.

Looking ahead, how might these programs evolve, and how will their enhanced capacities influence the future of architectural practice?

Generative computational tools are evolving quickly, however, there are certain significant thresholds that they cannot yet pass. For instance, transitioning from two-dimensional images to three-dimensional spaces is exponentially more complicated and, therefore, incredibly expensive computationally. But these problems can and will be solved, so the question remains: how can we use these tools to redefine architectural practice, and how can we be more ambitious about the scale of the challenges we choose to confront? It’s vital that we work collectively and use these tools to confront the planetary pressures facing us.