Inspiration
We were inspired by the prevalence of unconscious bias in everyday language and the need for a tool to promote more inclusive communication.
What it does
Our Chrome extension, Unbiasly, detects and corrects biased language in real time. It analyzes text and speech, identifies biases in various categories, and suggests more neutral alternatives.
How we built it
We leveraged advanced trained machine learning models. We also used speech recognition models to get the transcripts and detect biased language. The extension's frontend is built using JavaScript, while the backend utilizes Python and a robust API to process user input and return suggestions.
Challenges we ran into
One of the challenges was to identify the problem areas of occurrences of social bias environment and come up with a solution to mitigate it as much as possible. Additionally, there was pressure to incorporate a multitude of features in a short period.
Accomplishments that we're proud of
We're proud to have developed a tool that has the potential to significantly impact how people communicate online. Our extension provides a practical solution to a complex problem and promotes a more inclusive digital environment.
What we learned
Through this project, we gained valuable insights into the ethical implications of AI. We learned the importance of continuous learning and improvement, as well as the need to consider user experience and accessibility. It was an amazing opportunity to tackle new technical challenges.
What's next for Unbiasly - AI-powered Chrome extension
We plan to expand Unbiasly's capabilities by incorporating more sophisticated language models and expanding its coverage to a wider range of biases. We also aim to integrate the extension with popular productivity tools and social media platforms to maximize its impact.
Built With
- ai
- flair
- javascript
- machine-learning
- openai
- python
Log in or sign up for Devpost to join the conversation.