AI Society Tracker

Strengths:

- Novel Idea: The concept of a real-time AI societal impact tracker is innovative and addresses a critical gap in AI safety and policy.

-Relevance: The paper highlights the importance of tracking AI’s societal impact, particularly in areas like job markets, ethics, and policymaking.

-Groundwork: A strong foundation for future work, including identifying data sources and outlining the need for such a tool.

Areas for Improvement:

-Methodology and Documentation: The paper lacks a methodology section and detailed documentation (due to time constraints), making it difficult to evaluate the technical implementation. Include a clear explanation of how the tracker works and provide access to the codebase.

-Results and Insights: The results section is missing due to time constraints, and the app does not provide enough information to interpret the data. Include a detailed analysis of the findings and their implications.

-Limitations and Risks: The paper does not thoroughly analyze the limitations of the tracker itself or the potential risks of implementing such a tool. Expand on these aspects to strengthen the analysis.

-Scalability and Generalization: Demonstrate how the tracker could be scaled or generalized to broader applications. Include a discussion of future steps and scalability.

Suggestions for Future Work:

-Develop a detailed methodology and document the technical implementation, including access to the codebase.

-Conduct a thorough analysis of the results and their implications for AI safety and policy.

-Explore mitigation strategies for the risks and limitations.

-Investigate the tracker’s performance in diverse contexts and expand the data collection process.

The submission draws upon quite a few novel ideas for their 'AI Societal Impact Tracker' and should be appreciated for their innovative idea. The paper was also well written, and uses formal language in a clear, presentable manner. However, in order to strengthen the submission, greater usage of literature should have been applied. Unfortunately the lack of depth in the sources means that there is not enough literature as a foundation. There is plenty of existing research that surrounds the need to examine the impact of AI on society, therefore it would be recommended that in the future, the submission draws upon more research and statistics. The paper is also incomplete, which means there is less discussion and analysis that can be assessed for the submission. Overall a great effort.

Hi Abigail, I wanted to say that I appreciate the fact that you decided to submit your project despite it not being finished. Overall, I think the idea is great and very practically relevant, and something that you should definitely pursue if you’re passionate about. As to the project itself, you could use what you’ve written so far as a ‘call for action’ type of blogpost, or develop it further and publish as a peer-reviewed ‘perspective’ article. We can also help you further, if you want to build the actual platform.