# 20240311 Summary: GAICA Hackathon

## Participants:
- **Learning Lab (LL):** Madeleine Woods
- **STEM Leadership:** Greg and Eske
- **GAICAs:** Suvin, Nene (STEM GAICAs), Gonzalo, Indra (LL GAICAs), Emmanuel (absent)

## MVPs (Minimum Viable Products):
- **Student-Facing Slackbot:** Development by Suvin for a PHYSICS15b support tool.
- **Live Lecture Tools:** A set of tools, including student polling,lecture summarization, and resource document creation, run from Google Colab and managed by Nene.
- **Colab Autograder:** Created by Madeleine Woods (Learning Lab) w/ Greg (STEM Leadership) for automated grading systems.

## Timeline Goals:
### 1-Week Sprint:
- **Student-Facing Slackbot** for PHYSICS15b (Suvin)
- **Colab Autograder:** initial prototype (Madeleine + Greg)

### End of Academic Year:
- **Live Lecture Tools:** stack in colab (Nene)
- **Student-Facing Slackbot** evolution with added utilities, potentially faculty-facing features (Suvin + Emmanuel)

## Action Items:

- Database creation and management through AWS
- Security and privacy considerations taken into account.
- Slackbot deployment to 15b students by March 18th
- Greg and Madeleine meet to discuss Canvas APIs and autograding prototype
- Advanced Colab version roll-out for Nene and Suvin.

## Key Discussions:

* Security and Privacy:
- Slack implementation to include secure login, ensuring Harvard-only access.
- Discussion of privacy concerns and data handling similar to Canvas policies.
- Emphasis on faculty-facing tools for structuring learning and avoiding over-reliance on model output.
* Use of AWS for secure, scalable database solutions.
- Integration with Harvard's existing systems and compliance with security classifications (e.g., L3 protection).
- Structuring AI tools to enhance critical thinking and learning, rather than shortcutting the process.
- Considerations for the humanistic application of AI tools in pedagogy.
- Balancing the adoption of new teaching tools with traditional faculty workflows and concerns.