These discussions are for anyone interested in how to operationalize the AI ethics principle of transparency, especially via documentation of machine learning systems (similar to a nutrition label on food or safety data sheet for chemicals).
Documentation should be considered both as an artifact (the nutrition label) and a process (the procedures and actions that leads a system of humans to create the nutrition label for each component within ML systems).
You will find topics on:
- The latest academic research on how documentation can improve transparency
- Open questions on most impactful to disclose
- Stories of what processes people are attempting
- Challenges and victories in applying the research
- and more!
Please visit this forum to learn and contribute to this promising way to decrease bias and build more just ML systems.
This forum is an instance of Discourse. If this is your first Discourse forum, see the Discourse New User guide for instructions on how to get started and useful tips and tricks.