Themes and Insights

During my conversations with Raina, the AI (Artificial Intelligence) Chatbot for MagicSchool AI, I received well-organized lists with brief descriptions of key themes in Raina's responses. As we have explored throughout some of our courses and this week, Artificial Intelligence has the potential to impact education through data analytics and instantly be able to absorb and compile information from a vast number of resources sources.   The assignment from this week serves as an example of AI’s ability to do so.     

The amount of information provided was at first overwhelming. However, after examining Raina's responses on ethical considerations regarding data privacy and algorithmic bias, four main themes emerged:   

• Ethical practices and uses of AI   

• Addressing algorithmic bias  

• Ensuring data privacy and security  

• Engaging and educating all stakeholders involved in AI in ePortfolio development.  

In terms of addressing the digital divide to ensure equity, inclusivity, and accessibility, three key ideas that stood out to me were:   

• Collaborating and engaging with the community to tackle digital divide challenges  

• Designing with the user in mind to promote accessibility for diverse needs  

• Ongoing evaluation and improvement guided by data and feedback.  

Some of the information was not new information to me and although the extensive lists provided for the initial three questions posed were overwhelming, they were beneficial in highlighting points and ideas I had not previously considered before engaging Raina in an AI conversation. I was already aware that one of the primary ethical considerations of the use of AI is the potential for bias and discrimination (Khan, 2023), Yet, I had not given sufficient priority to the concepts of data minimization and establishing procedures for data retention and deletion or having an incident response plan in place.  

With the first and second conversations that focused on data privacy and algorithmic bias, I found that when I asked direct follow-up questions like “Can you provide examples of tools or techniques that can help safeguard student data in ePortfolios using AI?” and “What are some bias detection tools?”  Raina provided me with more specific and tangible tools, programs, and strategies to consider like data encryption, anonymization software, and multi-factor authentication. Raina also listed specific tools and programs to consider when I ask, “What are some bias detection tools?”  Because the initial question I asked was broader, I was given broad suggestions. Probing with more direct questions yielded specific tools and programs that I and/or a team could consider.  However, I would need to consider if these suggestions that Raina provided are high-quality tools and programs and I would need to further research how reliable and valid these programs are and read reviews from customers. This would not be an aspect that I think would be left up to AI to decide for me, considering the potential for algorithmic bias, and mitigating issues related to the digital divide.  

In a research study conducted by Syzdykova et al. (2021), they emphasize the benefits of developing ePortfolios at the Higher Education level,   

The advantages of an e-portfolio as an assessment tool are based on the fact that it encourages lifelong learning by making it possible to collect, monitor, and study student experiences. E-portfolio helps teachers and students improve meta-cognition, convey the true concept or new information, and use data analysis (p. 129).     

In their research article, Syzdykova (2021) et al. highlight the importance of additional research on e-portfolios, particularly in creating a teacher manual for evaluating student portfolios. Such tools can enhance teacher guidance to students and aid learning, particularly in online settings. Syzdykova et al. (2021) suggest a future direction involving AI-based systems that could offer personalized recommendations and tasks for students to enhance their professional profiles based on self-analysis data and teacher feedback (p. 131). Thus, it is an imperative AI-driven platform used to develop ePortfolios needs to address the vast number of ethical considerations about data privacy, and algorithmic bias, and mitigating the digital divide.