2023-08-11 Meeting notes

Date

10 Aug 2023

Attendees

Lawrence Eribarne

Goals

Discussion items

Time	Item	Who	Notes
	Training Needs	All	Faculty and Staff Familiarity with Generative AI (GAI): Most faculty and staff are not well-versed with generative AI. There's a pressing need to educate them quickly. Meeting Dynamics: Adam highlights the challenges of large meetings where participants might not feel comfortable expressing ignorance. Logan emphasizes that faculty skepticism might stem from the summer season and the need for individual sign-ups for personal ccounts. Need for Practical Support: Tamara stresses the need for concrete strategies, practical advice, and prompt structures for both faculty and students. Generative AI Training: Discussions revolved around different training sessions, the role of Bok, and feedback from pedagogy fellows. Honor Council Issues: Rebecca mentions challenges arising from misuse or misunderstanding of AI in the honor council. TA's Familiarity with AI: Greg's survey reveals mixed experiences among TAs with AI but a majority seems excited. Training Tools: The potential for various tools like Slack channels, videos, and HUIT tools were discussed, along with concerns around privacy and data storage. Faculty Staff: Guidance / training needed for TAs and TFs. Bok: will do onesession within fall teaching conference. This week did bootcamp for 27 pedagogy fellows and they were on the very positive, Mandatory Training: Crhis suggested training be mandatory before access is provisioned.
	AI Sandbox Demo	Ventz	Integration with Canvas: Jefferson discusses ongoing work to integrate GAI tools with the Canvas platform. Tool Development and Features: Various aspects of tool development, including scaling, data storage, and user experience, were explored. Future Vision: Jefferson suggests that success would be retiring current tools as more advanced enterprise tools become available. Features: Adam noted that submitting link to chat history is a requirement. Features: Chris pointed out there is one to many and many to one use cases (students across classes, and faculty across classes). Features: Rebecca pointed out that focus on data upload and code interpretor are the correct priorities for feature dev. Requests: Jeffeson pointed out tool will be provisioned on a request basis and capacity review. Future state will be to retire this tool and leverage enterprise options.

Action items

Logan McCarty Send a follow-up message to faculty with a recording of the session and post on the GAI webpage.
Gregory Kestin and Logan McCarty Ensure every instructor has an informed choice on syllabus policy regarding GAI. Course by course review - beginning with largest courses. Set prioritization for provisioning assuming limited capacity.
Rebecca Nesson Establish a Slack channel with VPAL dedicated to GAI instructional methodology. Propose making videos with undergrads, based on a list of topics. VPAL level Slack channel to be set up. Clarify training expectations and possible make-up sessions.
Rebecca Nesson Gather and/or record quick training or videos on uses. Leverage support of undergrads.
Lawrence Eribarne and Ventz Petkov Create chart for pros/cons of each model.
Lawrence Eribarne get GAI STEM team access to HUIT sandbox
Christopher Stubbs Issue a Science + SEAS information bulletin to faculty with recordings, updates, and emphasis on retooling.
Christopher Stubbs Check with Amanda on providing universal expectations for TA / TF appointments, maybe with OUB / GAS (sp?)
Lawrence Eribarne Report back on communication, training, and onboarding processes that emerge from the communications and service management GAI efforts
Ventz Petkov Report back on use cases for many to one or one to many.
Christopher Stubbs Create draft of uses as training material.