I made ThoughtDAG – LLM as an editable graph, wires are the context (github.com)

ThoughtDAG has launched a novel tool that transforms conversations with large language models (LLMs) into editable thought graphs, allowing users to visually map their ideas and interactions. This intuitive platform, which requires no installation or signup to access an example canvas, emphasizes a human-in-the-loop approach where users control the context and information the model utilizes. By literally “wiring” context into nodes, users can easily refine and manage their interactions with LLMs, resulting in cleaner, more relevant responses while avoiding extraneous noise in their queries. This innovation is significant for the AI/ML community as it enhances user engagement with LLMs, allowing for a more structured and dynamic way to interact with AI-generated content. Notably, ThoughtDAG supports various models, enabling seamless integration with OpenAI and other endpoints while preserving user privacy through local-first design. The tool facilitates collaborative and iterative exploration of ideas, with features such as real-time updates to answer relevance and backward-compatible formatting for permanent accessibility. This could lead to new methodologies in knowledge management and cognitive exploration within AI applications.