An AI knowledge twin is a digital avatar that contains the knowledge and expertise of its "owner" – a real human being. The AI twin's knowledge base includes the owner's scientific publications, teaching materials, and additional information about the owner and their experience.
Lecturers P. Jurčys and G. Strikaitė-Latušinskaja from Vilnius University Faculty of Law have decided to employ their AI Knowledge Avatars "Paul AI" and "Goda AI" as tools to assist in teaching a course on "Data Protection Law: A Human-Centric Approach."
Scholars have discovered that "Paul AI" and "Goda AI" offer students a new way to engage with their professors and access information by asking questions and discussing accumulated knowledge. Specifically, the knowledge base of "Paul AI" includes information on intellectual property rights, copyright, data protection, and innovation. On the other hand, the knowledge base of ‘Goda AI’ includes information on law enforcement processes, court proceedings, and the application of data and AI technologies in the public sector.
"This new form of interaction with the lecturers' knowledge through their AI knowledge twins seems like an excellent opportunity to provide students with additional ways to engage with the course material in the 21st century. This seems especially important considering that students' attention spans in the world of mobile applications are constantly shortening," says Dr P. Jurčys.
In addition to "Paul AI" and "Goda AI", lecturers encouraged students to use all other generative AI tools, including #ChatGPT, Inflection AI's Pi, Anthropic's Claude, Perplexity's search engine, you.com, etc.
The objectives of the course
The intensive course was designed to integrate theoretical knowledge with the practical application of knowledge and the use of AI tools. The first four lectures on "Data Privacy Law" focused on the principles of a new, human-centric approach to personal data. These lectures aimed to highlight new data usage models that are intended to create new ways of using personal data for the benefit of individuals.
During the preparation for the lectures, students were encouraged not only to read the necessary literature but also complete specific practical tasks using "Paul AI", "Goda AI", and other AI tools.
During subsequent lectures, students were divided into groups, tasked with selecting topics related to data usage, identifying specific problems, and preparing presentations on solving those problems. Students were encouraged to utilise any AI tools of their choice.
"These group projects and the collaboration with AI tools provided an excellent experience. Students had the opportunity to explore the issues that interested them most and present specific solutions to socially significant problems. They created potential business models and even mobile phone apps, despite the fact that it was a course taught at a law school. Students also developed legislative proposals," points out G. Strikatė-Latušinskaja.
Regarding the lecturers, the change in the teaching format provided students with increased freedom, expanded opportunities to showcase their creativity, and facilitated collaborative learning not just on the implementation of legal principles in real-world scenarios but also on the utilisation of AI tools in prospective legal endeavours.
The exam, using AI tools, and lessons learnt
Another challenge in this project at Vilnius University Law Faculty was determining how to assess students' knowledge. What would an exam look like if students were allowed and encouraged to use AI tools during the exam?
To address these issues, Dr P. Jurčys and G. Strikaitė-Latušinskaja sought input from educators and EduTech experts worldwide, including the USA, Japan, Denmark, and the Netherlands, to create the exam assignments. The aim was to evaluate the understanding acquired from lectures on the core concepts of data protection law and their real-world implementation.
The integration of the lecturers' personal AI knowledge twins – "Paul AI" and "Goda AI" –demonstrated the potential of AI to enhance the educational experience by providing students with personalised, round-the-clock assistance and comprehensive feedback.
"With the assistance of our AI knowledge twins, we, the instructors, were able to create a more dynamic and interactive learning environment that better meets the expectations of students in today's technological landscape," says Dr P. Jurčys.
Similarly, the course emphasised the significance of combining theoretical understanding with real-world application. It motivated students to tackle practical challenges using AI tools, which in turn, demanded critical thinking, even when utilising AI, and fostered essential digital literacy skills crucial for the evolving legal industry.
The experience revealed the necessity to experiment and explore positive ways of applying AI tools. The utilisation of AI knowledge twins underscored the importance of continuing the discussion on implementing AI in education.
"While AI tools offer tremendous potential, their use raises important questions about academic integrity, data privacy, and the role of human oversight in the use of specific AI tools. As AI technologies evolve rapidly, it is crucial to ensure their responsible and ethical use," believes G. Strikaitė-Latušinskaja.
