Team members: Michael Anti, Ming Fan
Summary: This proposal presents the use of an AI Agent to test improvements to real-world Voting Protocol with the goal of enhancing the quality and acceptance of voting outputs in areas such as social media, markets, and democratic elections. Current Voting Protocol, despite being widely used in a variety of decision-making processes, still suffers from obvious problems such as the emergence of populist leaders, fake news, and cyberbullying. Improvements are centered on optimizing voting timing and computational modules through simulation testing.
Specific research processes, including discussions with experts, comparison of social media algorithms, and AI simulation of the voting process. In addition, the effectiveness of these improvements will be tested through programming simulations and online workshops, with the ultimate goal of sharing the research results and tools through open source in hopes of providing references and tools for the improvement of global democries and social media.
1) What is the existing target protocol you are hoping to improve or enhance?
The target protocol is Voting Protocol. Voting Protocol is the real engine behind stock markets, sales markets, democratic elections, social media, and meme communication. DEFINITION: Voting Protocol is the machanism of ordinary people to influence more people by selecting, either online or offline, through their own favor and recommendation, a peer or a work which is perceived to be prominent from a community-wide perspective.
In fact, many real-world heroes (leaders and celebrities) have gained this status as a outcome of Voting Protocol. And because Voting Protocol is so effective at resisting boredom and selecting the exceptional from the mediocre, markets, democracy, and social media have always had a tremendous efficiency and competitive advantage. Even the current successful training of LLM is due to RLHF - a feedback process based on Voting Protocol.
2) What is the core idea or insight about potential improvement you want to pursue?
But real-world voting protocols, both online and offline, are not very satisfying.
First, the mechanism by which voting protocols are introduced into the information process has not been studied enough.The TikTok algorithm has a special timing for the introduction of user voting computing, so much so that its recommendations are considerably better than those of other social media, suggesting that the Timing of what to put Voting Protocol in is critical to the outcome. Also, while the Chinese have the largest social media in the world and practice Voting Protocol on their cellphones every day, this has not in any way contributed to the emergence of democratic elections in Chinese society. It is not particularly clear to us why it is easy to accept one kind of voting but difficult to accept another.
Second, the results of real-world voting also make for mixed results. Markets can fluctuate wildly, elections can elect populist leaders, and social media users are victimized by shoddy content, fake news, and cyberbullying. In order to produce quality results, Voting Protocol needs to be improved.
From a system perspective, a Voting Protocol is a feedback system, except that all feedback signals are returned to the output for individual inputters only after total sorting and other computing. In order to improve the output, it is necessary to optimize the whole system by adding new computational modules, feedback loops at appropriate locations.
Therefore, we hope to find the right timing to introduce the Voting Protocol in the system, and the computational modules to improve the quality of the voting results by studying the the real-world voting protocol, and improving the computational modules of it.
3) What is your discovery methodology for investigating the current state of the target protocol?
We will visit and discuss with experts in political science to prototype the existing democratic voting process; we will work with algorithmic engineers to compare the efficiency of various current social media for generating quality content and make algorithmic comparisons; and we will work with scholars of Internet Politics and Political Psychology to compare the similarities and differences between political elections and social media at the protocol level. We will also make a Case Study database of the advantages and disadvantages of real-world Voting Protocol.
4) In what form will you prototype your improvement idea?
We do improvement experiments based on the above prototype using AI Agent technology to simulate the human voting process. Generative AI gives us the ability to simulate the judgment of an ordinary person, and AI Agent can simulate prototypes of complex voting protocols.
5) How will you field-test your improvement idea?
By programming simulations, comparing the quality of various modular improvements to voting system outputs, and convening an online Workshop for peer review. At the same time, real-world cases are analyzed and to improve the prototype.
6) Who will be able to judge the quality of your output?
Social media, political organizations, and political scientists are able to review and adopt test our improved prototypes, such as AI experts, TikTok algorithmists, Twitter techies, Berkman Klein Center for Internet & Society and Nieman Lab, etc.
7) How will you publish and evangelize your improvement idea?
The entire simulation test code, dataset, cases, and conclusions will be open sourced on GitHub.
8) What is the success vision for your idea
We think about voting protocols over and over again because of a paradox that happens in my home country: we have the largest (if not the best) social media in the world based on Voting Protocol (Tiktok, Weibo, WeChat), but we don’t have a single democratic election. Therefore, there must be some unrevealed secrets hidden in the protocol level.
If our research is successful, it will not only improve social media globally and democratic voting in the West, but also bring some guiding lights to the freedom of Chinese people.