Our project aims to explore the application of ChatGPT, LLama, and other LLM models within the context of Cosmos community governance. We propose using the Cosmos forum discussion dataset to train a bot that can make informed decisions on governance proposals. Furthermore, we can imbue the bot with the personality of historical figures like Abraham Lincoln, Karl Marx, etc., by incorporating their public writings, which reflect their values, preferences, and political-economic thought processes.
By focusing on intellectual knowledge as the basis for decision-making, we strive to remove greed and self-interest from the equation, paving the way for a more objective governance system. In the Cosmos context, this approach could help address the issue of validator bribery, which has been a significant challenge thus far. We believe that our research has the potential to make a fundamental contribution to large-scale human governance systems.
