Fact Checker

• 

  Side Panel Explainer

Inspiration

The inspiration for this project came from the ever-growing volume of information shared on social media, where misinformation can spread rapidly and influence public opinion in significant ways.
I wanted to create a tool that would help users critically assess the validity of the content they consume online, empowering them to make more informed decisions. Leveraging generative AI for fact-checking felt like the perfect solution, allowing for nuanced and context-aware evaluations of social media content.

What It Does

Fact Checker is a browser extension that uses generative AI to assess the validity of social media posts.

How I Built It

I built Fact Checker using the new generative AI API.
The core of the extension is a prompt-based interaction with the generative AI model, designed to accurately evaluate social media content.
The extension gathers information from post, processes the data, and communicates with the generative AI to get a well-rounded, factual assessment.
The extension was developed using TypeScript and integrates smoothly with the browser's environment, with the generative AI API providing the fact-checking intelligence.

Challenges I Ran Into

One of the biggest challenges I faced was fine-tuning the prompts. Getting generative AI to deliver reliable and fact-based assessments required a lot of iterations and experimentation. Finding the right way to ask the AI for nuanced analysis, while avoiding false positives for satire or opinions, took considerable effort. Additionally, staying within the token limit of the current generative AI API restricted the length and depth of analysis we could perform on individual posts.

Accomplishments That We're Proud Of

I'm proud that Fact Checker can provide a meaningful and concise evaluation of social media posts. Successfully combining generative AI with practical, real-time analysis of social media content is a big step towards mitigating the spread of misinformation. I'm also proud of creating a prompt structure that balances depth of analysis with the limitations of the current generative AI capabilities.

What I Learned

Through this project, I learned a great deal about prompt engineering, especially in the context of ensuring reliable and factual responses from generative AI. I also gained experience in integrating generative AI capabilities into a browser extension, which involved tackling issues around response speed, token limits, and contextual understanding. Most importantly, I learned the importance of iteration when working with AI—the value of testing, refining, and repeating.

What's Next for Fact Checker

The next step for Fact Checker is to expand its scope. As the generative AI API becomes more robust and the token limit increases, I'll be able to extend the fact-checking capabilities beyond short social media posts to full-length articles. This would greatly enhance the value of the extension, providing more comprehensive analysis and helping users navigate the vast landscape of online information with confidence.

📝 Note:

• For this extension to function properly, the ai.languageModel API must be enabled in your browser.

Built With

• languagemodel-api
• react
• typescript
• webpack