Setting back to ETC
It is the first week of school. Unlike other semesters, we will be connecting from home for the first 2 weeks due to Carnegie Mellon University (CMU) regulations to prevent major Covid spread.
Introduction
Despite being remote, it is still nice to finally see each other after the winter holidays. Say “hi” to our team on zoom!
With our first meeting, we decided on the roles everyone will play on this team and shared with everyone’s strength and weaknesses, including past experiences. Here’s a rough breakdown of our roles in the team.
Equipment Loaning
With the start of the semester, we also started the process of loaning resources such as Wacom tablets from the Entertainment Technology Center (ETC) so that we can get started.
Name Brainstorm
Throughout the first week, we have been trying to think of a name to fit our project on biasness in AI. After much discussion, we finally decided on the name AIAuditors. We came up with this name as we are not building AI models, we are hoping to act as auditors to help discover the biasness and errors formed by AI models.
Readings before Client Meeting
After our initial meet up among ourselves and with our faculty advisor, Mike, we have also contacted with our client from CMU Human Computer Interaction Institute (HCII) to schedule a time to learn more about the project. In return, our client, Ken, has kindly sent us a few resources, aka 3 research papers, to read before the meeting.
It was kind of a struggle to read research papers on week 1 of school. Luckily, we are a team. We managed to collate the insights we gained from the readings into a google document to be shared with everyone.
First Client Meeting Summary
Finally, we had our first client meeting on Friday of the week. Unlike the usual sharing through a powerpoint by the client, it was more of a open question and answer question for us to clarify whatever doubts we had about the project. Everyone had an insightful discussion and it lasted for about an hour.
Through the meeting, we defined a problem requirement statement which we hope to solve through this project.
To summarize: We will create a fun game, built on top of a single model concept (api), that attracts public users to play, have fun while collecting data in the background to discover questionable judgments in the model.
Here’s also a list of things which our client hopes to achieve (ideally):
- Both individual and groups (collaborative & competing)
- Game that allows researchers to make meaningful research
- Async or sync prompts throughout the game at various points
- A feature for users to discuss or resolve disagreements between one another
As you can see, this is only week 1 and we already received such a broad vision moving forward. Stay tuned to see how we, as strong resilient ETC students, handle this project.
If you are interested, I have added a summary of the meeting minutes below:
Comments are closed