Using AI Assistants for PBL

I have not shared research on the potential of generative AI in higher ed for several weeks (Aug 30, 2024). I do receive a wide variety of questions from colleagues on a weekly basis on how they might integrate GenAI in a purposeful, meaningful and safe way into their teaching and learning. So, this week, I would like to share resources (perhaps not specifically SoTL) on one way to think about adding GenAI for those of you offering project-based learning (PBL) or anytime of group work.

Recall that PBL has several advantages:

Students deal with real-life problems, which helps them understand how theories can be applied to solve practical problems. This results in long-term retention and can lead to better learning experiences (Lehmann, 2008).
Students become more motivated when assessments are perceived as authentic. PBL will improve skills such as professional identity and awareness, communication skills, and their employability prospects (Sotiriadou et al., 2022).
Projects are formative assessments which provide timely feedback for improvement that improves performance (Collett, 2008).

Most likely most, if not all of you have heard and been a part of AI Meeting Assistants. These have become commercialized with the intent of monetizing video conference conversations (here is a link to the Zoom AI Assistant with start up videos). An AI meeting assistant can offload tasks such as scheduling and rescheduling, meeting transcription and translation (creating a more inclusive environment for non-native speakers), note taking, and creation of meeting recaps and summaries. Therefore the hope is that participants can stay focused on the ongoing discussion. Although these are indirect measures (survey), Zoom has collected data that represents how AI Meeting Assistants can save time.

Many of us have been encouraging students to use various recording tools when they meet with their project teams. Tools that could capture student conversations include “old-fashioned” digital audio recorders, voice recording apps on mobile phones, voice to text function on Google Documents, and more. The upgrade that an AI Meeting Assistant can add is the ability to summarize the discussion, create action items, add deadlines to calendars, create follow-up questions, track timelines, etc. All of this information can be shared with the instructor, which can provide documentation of accountability, which can be troublesome with group work.

Additional features of AI Meeting Assistants for PBL could include:

In-progress meeting catch-up. The AI assistant provides an up-to-the-moment recap for participants who join after a meeting has started.
Email recaps. The AI assistant captures keywords and action items and then distributes a meeting recap to participants via email.
Post-meeting queries. The AI assistant provides answers to natural-language prompts, such as, "Give me the top three takeaways for my team."
Qualitative pattern analysis. The Assistant can analyze text for keyword extraction, sentiment analysis, topic modeling, text analysis and summarization, comparison, thematic analysis, and clustering.

If you are interested in using AI Meeting Assistants, here are a few recommendations:

Otter for asking questions about your meetings [I have been using this for quite a while]
Gemini for Google Workspace, built into Gmail, Docs, Sheets with enterprise security
Zoom AI companion identifies important information in a convenient side panel
Fireflies for collaboration and topic tracking
Avoma for conversation analytics
Equal Time for inclusive meetings
Fathom for a free option

References

Gratchev, I. (2023). Replacing exams with project-based assessment: Analysis of students’ performance and experience. Education Sciences, 13(4), 408. https://doi.org/10.3390/educsci13040408