AnnotateFWD

• 

  Landing Page

• 

  Sign Up Page

• 

  Chatting Page

• 

  Low Fidelity Sketch

Inspiration

Many studies surrounding the accuracy of AI questions the passing of biased or incorrect information as completely true. This has to do with a distinct lack of equitable involvement in the training and verification of LLM and AI models. In particular, groups of people who are more likely to be misrepresented by these models are often not provided opportunities to help rectify them. This is a major factor in the caution users must take when using the information received from these outputs and may contribute to the stereotyping of marginalized groups. This inspired our problem statement: How can we design a means of verifying the outputs of LLM and AI models such that the voices of marginalized groups are properly incorporated?

What it does

Our web application allows for users to register and receive training on how to give feedback on AI and LLM outputs on topics related to their fields. The way AI is incorporated is having users prompt the model, receiving a response, and rating it based on its correctness and fairness. Afterwards, the users can submit the feedback and it will be reviewed by moderators to ensure the service's proper usage.

How we built it

We first discussed about current ethical concerns surrounding the use of AI generated content and came down to the idea of the lack of reliability, transparent, and accessibility for these generations. Afterwards, we developed Ethical Matrixes to list down the stakeholders and any concerns and values they would have. Then, we brainstormed different use cases and job stories to ensure that we would be able to consider realistically how the stakeholders would use and be effected by the design of the product. Then, we created a low-fidelity prototype through sketches and when we were happy with the ideas, we moved onto Figma to create a high-fidelity prototype that would also be iteractive.

Challenges we ran into

Both of us are have little to no experience working with Figma, and with only two members we also had to focus on managing our time well throughout this hackathon. Through the use researching how to use Figma, we were able to overcome this and ensure that we would be able to get the work done to make it look complete rather than overreaching our limits.

Accomplishments that we're proud of

We are both proud of being able to create a functional prototype for this project considering both of our lack of experience in prototype and missing team member.

What we learned

It was a new experience for the both of us but we have both learned various new skills surrounding UI/UX design.

What's next for AnnotateFWD

If there is an opportunity, we would develop this product into an actual website with a full stack development team and also connect with LLM models or use their API's to do their testing and give user feedback.

Built With

• figma