EXPERIMENT II: ATTEA

ATTEA is a chatbot that is part of the community for the Theory: Text Analysis course in the Master's in Architecture program at the Universidad de los Andes, taught jointly with Professor Ingrid Quintana. Its knowledge base is built on the work produced by students taking the course. Its expertise is limited to the texts analyzed throughout the course, as well as the debates, seminars, and writings derived from said analysis.    

To obtain sufficient data to train ATTEA, the students read architectural texts and compared opinions, findings, and inferences, reaching conclusions that led to valuable lessons. This was achieved by reviewing and analyzing different types of documents on the work of influential figures in the field of architecture at the national and international levels.   

As a deliverable, the students produced a series of texts, including reviews and opinion pieces, as well as audio recordings with peer assessments of their peers' work. These inputs were used to form the dataset used to train ATTEA. Subsequently, the students refined the chatbot by collecting interactions with the interface, identifying errors in its responses, and making the corresponding corrections.

. 

In addition to being able to interact with ATTEA, below you can read the reviews generated by ATTEA, requested through prompting from each of the course students, and opinion pieces about their experiences integrating generative AI into their learning processes and text creation during the course. Additionally, you'll find a podcast generated from the reviews written by ATTEA.