From image to object

Architecture imagined on paper, reinterpreted using artificial intelligence and finally transformed into a real object. Beyond the Frame returns on Monday 9 November at 6.00 pm with a new workshop led by Ilaria La Manna, aimed at teachers in upper secondary schools. The event forms part of PariPasso, the Rizzoli Education project dedicated to educational inclusion, and continues the programme launched on 19 October with the first session on AI, drawing and architecture. In the previous webinar, the work had begun long before the prompt: observation, drawing, geometry, proportions and functions had been used to understand architectural structures such as the Pantheon, the Palazzo della Civiltà Italiana and the MAXXI, before attempting to transform them through the design question ‘What if…?’. A building could change its function, architectural language, materials or relationship with its context, using AI as a tool for research and reinterpretation [see the article Beyond the Frame]. On 9 November, the programme picks up again from those very images to tackle the next step: how do you transform an AI-generated image into something that can actually be produced?

When the image must become buildable
One of the first steps is to look at what the AI has generated with fresh eyes. An image can be evocative and, at the same time, impossible to transform into a model. In the activities prepared by Ilaria La Manna, even the prompt takes on a new function. It is no longer merely for imagining: it must help make the project more realistic and structurally plausible. In one of the examples, the AI is asked to avoid modules suspended in space and to stack them in a coherent manner, conceiving the architecture as a real exhibition building. In another step, people are removed from the image to facilitate subsequent three-dimensional processing. It is a detail that encapsulates the essence of the workshop: generating an image is only the beginning.

From image to 3D model
The next step takes us from two to three dimensions. During the workshop, AI-based tools such as Meshy, Tripo 3D and 3D Valley are explored to transform an image into an initial three-dimensional model. However, even the 3D model is not yet ready to become a physical object. It needs to be exported, for example in .stl format, checked, any errors identified and corrected, the scale chosen, and the need for support structures assessed before preparing it for 3D printing. The slides guide participants step by step through this process, right up to the preparation of the print plate and the final printed object. The workshop thus brings the entire design cycle into focus: idea → image → model → verification → correction → production.

A process that trains students to design
It is precisely this sequence of steps that gives the workshop its strong educational value. Students do not merely produce an image using artificial intelligence: they must observe, interpret, choose, verify and correct. To arrive at an object that can actually be printed, they must in fact consider the coherence of the shapes, the proportions, the structure, the scale and the feasibility of the imagined solution. When the model contains an error, it must be identified and rectified. When an element cannot be produced, it must be modified. The error thus becomes part of the design process and an opportunity to refine one’s reasoning. For classes, this means linking together skills that are often practised separately: art appreciation, drawing, geometry, spatial representation, design, critical use of AI and digital fabrication. The transition from image to object also makes the consequences of one’s choices visible and helps to understand the gap between a plausible representation on screen and a solution that must also function in the physical world. The work can also be organised into small groups, with different tasks assigned: some analyse the artwork, others develop the concept, some construct and refine the prompts, others check the 3D model, and others document the transformations. In this way, the final product becomes the result of a shared process, in which subject-specific skills, creativity and the ability to collaborate all contribute to the same experience. 
It is also a practical way to develop a more informed use of artificial intelligence: the generated output is verified, modified and brought back into line with real-world constraints. Technology thus becomes a tool within a design process that remains guided by the students’ knowledge and decisions.

And it’s not just about 3D
The same image can also take a different path. The webinar demonstrates how to simplify the AI-generated output by transforming it into a minimalist two-dimensional representation, composed of just a few geometric elements and colours, suitable for subsequent graphic processing and digital fabrication. Drawing, art history, design and digital technologies thus come together in a single process, in which knowledge of the artwork remains the starting point and technology expands the possibilities for representing and transforming it.

Preparing for the workshop
The session is highly hands-on. To take part in the activities proposed during the live session, participants are invited to access the two workspaces in advance: Mentimeter, for interactive activities
Padlet, for uploading the images that will be used during the workshop
The programme also provides a PariPasso Toolkit to help teachers replicate the various stages of the experience in the classroom, from observing heritage to designing and bringing the idea to life.

Beyond the Frame | AI, Drawing and Architecture Workshop
REGISTER FOR THE WEBINAR
Monday 9 November 2026, 6.00 pm
with Ilaria La Manna

From paper to physical form, via artificial intelligence: a design idea is put to the test until it becomes something that can be observed, corrected, built and touched.