Celebrate Learning Week Workshop- May 8th by

fatemehsalehiankia — 2024-05-08 18:41:52 UTC

Here are some helpful resources on how to use advanced AI assistants and a conversation around the ethical implications of these emerging AI technologies:

Co-Intelligence: Living and Working with AI by Ethan Mollick
The ethics of advanced AI assistants by Iason Gabriel
AI Pedagogy Project by metaLAB at Harvard
Raji, I. D., & Dobbe, R. (2023). Concrete problems in AI safety, revisited. arXiv preprint arXiv:2401.10899.
Chicago
Shibani, A., Knight, S., Kitto, K., Karunanayake, A., & Shum, S. B. (2024). Untangling Critical Interaction with AI in Students Written Assessment. arXiv preprint arXiv:2404.06955.
Fleckenstein, J., Meyer, J., Jansen, T., Keller, S. D., Köller, O., & Möller, J. (2024). Do teachers spot AI? Evaluating the detectability of AI-generated texts among student essays. Computers and Education: Artificial Intelligence, 6, 100209.

Question for leadership, faculty

2024-05-08 19:35:40 UTC

How do we ensure that UBC community members (faculty, students and staff) have equitable access to LLMs that are developed and deployed to support teaching, learning and research, and evaluate that our approach is achieving this?

For students

2024-05-08 19:38:31 UTC

How can LLMs be used in a way to positively impact learning and how can the institution/instructors/peers discourage or prevent their unethical use

Instructor

2024-05-08 19:38:44 UTC

How do LLMs best supplement my students' learning without becoming the sole provider of information and solutions moving forward? How do we best incorporate these tools into the classroom while preserving (and teaching) critical thinking skills?

Questions for Instructors

2024-05-08 19:38:51 UTC

How might we support/measure pedagogical effectiveness? How might students’ learning be enhance through the use of AI? Are there frameworks for implementation that are working well in your area? What ethical considerations are you putting in place to protect learners and instructors academic output?

Question for instructors and Faculty leadership

2024-05-08 19:39:07 UTC

How will you measure the impact of the integration of Gen AI on student learning achievements? How will you measure the impact of changes you are introducing into the learning and assessment activities due to changes you make with these tools?

Question for every stakeholder

2024-05-08 19:40:24 UTC

How will the widespread use of LLMs change the values of the institution over time?

Question for Researchers

2024-05-08 19:40:32 UTC

How to make sure of the accuracy and reliability of the research findings?

How the LLMs take into account data privacy and ethical considerations? What criteria are they based on?

how to make sure there is the possibility of removing identifiers (in the background) once the computation/modelling is done?

Legal/Compliance

2024-05-08 19:40:55 UTC

How do we best protect our end-users at UBC from leaking sensitive information out into these LLMs? Do we look to licensing one such model, and then make it available across campus for "safe" LLM use?

Researchers

2024-05-08 19:41:03 UTC

With LLMs evolving so quickly and subject to user interactions with them, how to you: (1) explore parameters in the "black box" (2) ensure reproducibility (3) meet research ethical requirements such as the participants' ability to remove their data from use in your study at any time?

Question for Instructors

2024-05-08 19:41:57 UTC

How can instructors navigate academic dishonesty/plagiarism while supporting student use of LLMs?

On automation and confirmation bias

2024-05-08 19:42:19 UTC

Impact on student’s learning/ ethical consideration .

AI generative tools amplify the bias built in the data set they have used for learning. How can we measure this impact in student learning and importantly how can we train them to interrogate the tool properly to surface this bias and mitigate potential wrongdoing with a critical approach?

IT Infrastructure Support

2024-05-08 19:42:21 UTC

Building on a licensed UBC LLM, how do we build the infrastructure behind supporting and maintaining an LLMs use at UBC if we decide to forbid/deny access to Cloud-based LLMs?

Learning technology support

2024-05-08 19:43:05 UTC

How do you find a balance between faculty/students exploring this rapidly emerging space and institutional requirements for data protection and ethics?

University Leadership/Faculty

2024-05-08 19:43:44 UTC

How do we define (or redefine) ethical use of LLMs to promote learning while inhibiting mass plagiarism?

Sources of Student Data That Can Be Embedded

2024-05-08 20:11:15 UTC

-Canvas course clicks/engagement relative to other students in the course

-Module progress relative to other students in the course

-non-identifying demographic data

-quiz or assignment results compared to others --> Feedback on learning

-student desired goals for the course (self-identified)

-individualized learning profile (also self-identified by students) using a robust, research supported metric or tool

2024-05-08 20:12:46 UTC

Student data that can be embedded in the LLMs:

1. Information on how to identify learning skills and what tools or skills are required.

Information on previous study skills (time management, self motivation, etc. )
Our students identify challenges and experiences thy had prior to their UBC studies. If there is a way to integrate that data into this tool, students can get more personalized feedback from the AI coach

Question 2

2024-05-08 20:14:32 UTC

How do you feel you are doing in the course?
How are you actually doing in the course?
How important is this course for your learning goal/path
Is this your major?
What are you planning to do when you graduate?

Question 1 and 2

2024-05-08 20:18:54 UTC

Ask Justin. He'll be able to answer this both ;)

A recap of how they did in prior course

2024-05-08 20:19:25 UTC

When they start the current course and as the course continues, the student could be reminded of how they did in the prior course they took, what worked and what didn't so they avoid similar pitfalls.