June 9, 2024

Urban Planning & AI: GPT 4o as Architect

Sofia Malmsten

CEO & Architect

OpenAI’s latest release, GPT-4O, has sparked a flurry of discussions across social media about its potential to revolutionize the field of architecture. Enthusiasts and professionals alike are intrigued by the question: Can GPT-4O be an architect? What is needed for it to become one? And how will this development affect the role of the architect and traditional processes in our industry?

This article delves into GPT-4O’s impressive capabilities, the enhanced GPT experience tailored for personal assistance, and its transformative potential within architecture and urban planning.

Expanding Creative Horizons

One of the most compelling uses of GPT-4O in architecture is its ability to enhance creativity. Architects can leverage AI for text and images to generate innovative design ideas, pushing the boundaries of conventional thinking. The AI’s ability to provide unique and fresh perspectives can inspire architects to explore new concepts and design paradigms. However, the model you encounter in ChatGPT is a general model, and a new type of value comes into play with custom GPT models tailored to specific processes and domains.

When using ChatGPT for a task, it applies generalized knowledge and provides outputs that may be similar across different users. In contrast, a Custom GPT is fine-tuned for very specific tasks and operates in a bespoke, domain-specific setting to excel at precise functions. These custom models might be accessible to the public, exist within a single business, or be created for individual use.

Building on Existing Models

The custom GPT approach emphasizes creating efficient workflows, effective management systems, and connecting AI-supported processes to achieve real-world efficiency. In a recent YouTube video ((, landscape architecture is discussed, showcasing how GPT-4O can assist in creating intricate and imaginative designs. This highlights the AI’s potential not only as a creative tool but also as a collaborator in the design process.

Emphasizing this collaborative role of AI opens up a new technological era for product development. While the end client can certainly use AI in their daily work, the real transformation occurs when AI is integrated into a series of workflow tasks. This integration can significantly accelerate efficiency and development in the Architecture, Engineering, and Construction (AEC) industry.

By combining AI with structured workflows, we can streamline processes, reduce errors, and enhance overall productivity. This synergy not only saves time but also fosters innovation, enabling professionals to focus more on creative and strategic aspects of their projects. The potential for AI to revolutionize the AEC industry is immense, leading to smarter designs, quicker project completions, and ultimately, a more efficient and dynamic industry.

A practical example: Bringing in AI to a planning tool for architects

At Parametric Solutions, we are integrating AI to explore a wide range of improvements, assisting and enhancing existing workflows. We are not developing an AI tool from scratch; instead, we are creating a tool that is augmented by AI. Here are some examples of how AI will facilitate the architectural process in Hektar:

  1. Describing Scenarios
    As a result of the Vinnova project “Scaling up AI solutions” we have demonstrated how AI and LLM:s can describe and understand architectural qualities. By integrating LLMs and architectural 3d models we can now generate detailed descriptions of architectural scenarios. By analyzing architectural qualities, highlighting risks, identifying possibilities, and pinpointing areas for improvement, AI can help users quickly grasp the essence of a design. Try it out here:

AI describing the qualities, risks, and architectural aspects in a generated 3D volume in Hektar.
Experiment where AI describing the qualities, risks, and architectural aspects in a generated 3D volume in Hektar.

  1. Prompting Scenarios
    Soon, this capability might be enhanced, allowing users to describe their desired outcomes, which the AI will then translate into 3D models.
  2. Site Analysis
    AI can conduct comprehensive site analyses by examining images, topography, and sightlines. This assists architects in understanding the context and constraints of a site, enabling more informed and context-sensitive design decisions.
  3. Interpreting Planning Documents
    GPT-4O can interpret complex planning documents, making it easier for architects to navigate regulatory requirements and ensure compliance. This capability streamlines the design process, reducing the time spent on administrative tasks.
  4. WMS Services and GIS Collection
    By presenting relevant layers from WMS services and GIS collections, AI can provide architects with critical data about the site. This includes information on environmental factors, zoning regulations, and existing infrastructure, which are essential for informed decision-making.

The Future of AI in Architecture

By leveraging these capabilities, AI can significantly enhance the early-stage planning process in architecture, making it more efficient, accurate, and innovative.The Future of AI in ArchitectureTo sum up and revisit the question of whether GPT-4O can act as an architect and what would be needed: the integration of AI like GPT-4O in architecture and urban planning heralds a new era of innovation and efficiency. By enhancing creativity, streamlining workflows, and providing valuable insights, AI is set to transform how architects approach design and planning.However, it’s important to note that while AI can significantly augment the architectural process, it cannot replace the nuanced expertise and creativity of human architects. The architect’s role will evolve, focusing more on conceptual design, critical decision-making, and the integration of human-centric elements that AI cannot fully replicate. Architects will increasingly act as directors of technology, guiding AI to achieve the desired outcomes while infusing projects with their unique vision and understanding of human needs.

The possibilities for creating more sustainable, efficient, and imaginative spaces are boundless as we continue to explore and implement these technologies. GPT-4O is not just a tool; it is a catalyst for change in the architectural industry. By building on its capabilities, we can unlock new levels of creativity and efficiency, ultimately leading to better-designed spaces that meet the needs of our ever-evolving world.

As AI technology advances, it will enable architects to tackle complex challenges with unprecedented precision and creativity. The future of architecture will be shaped by the harmonious collaboration between human ingenuity and AI capabilities, resulting in spaces that are not only functional and beautiful but also sustainable and adaptive to the changing needs of society. This synergy will empower architects to push the boundaries of design and redefine what is possible in the built environment.