Abstract
We present the results of an early experiment with using generative artificial intelligence (GenAI) for creating digital three-dimensional (3D) reconstructions of archaeological remains, with GenAI also acting as an assistant to students learning cultural heritage. Although GenAI has been widely applied to textual learning and teaching for several years, 3D modeling offers a new and complex area for exploration. During a couse on East Mediterranean Archaeology in the fall of 2025, students are learning about the human past by creating a 3D model of a building from an ancient site in that region. Since archaeological remains are often found in a fragmentary state, real archaeologists need to use all forms of evidence to make a best estimation of what an original building looked like in the past. With such a reconstruction, we can better understand the use of space by people. This 3D assignment provides an opportunity for the students to experience this type of authentic archaeological research. During 3D modeling, students must make explicit decisions about how each wall, door, floor, and roof would look and be structured in the past. The act of modeling forces the student to think about every detail, since none can be ignored in the 3D model. We have facilitated learning through 3D modeling in many past archaeology courses. Now, we are experimenting with what happens if we add GenAI into this process, to act as an assistant for students to build the 3D models. Will this help them to reason about the reconstruction from the evidence, or will it create more confusion through GenAI’s lack of factual knowledge about the evidence? Will students still gain practical skills in using 3D modeling software, do they need to, or will they gain more time to focus on the underlying reasoning if they are no longer focused on the tool? We plan to report on the results of our experiment regarding these and many other potential questions.
Presenter(s) / Author(s)
