Reproduce “StorySage: Conversational Autobiography Writing Powered by a Multi-Agent Framework” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Shayan Talaei, Meijin Li, Kanu Grover, James Kent Hippler, Diyi Yang, Amin Saberi
StorySage introduces a multi-agent AI framework designed for conversational autobiography writing, helping users transform scattered personal memories into coherent narratives while prioritizing human agency. The system demonstrates substantially improved biography coverage and user satisfaction with the co-creative process compared to a state-of-the-art baseline.
Key insights
- Decomposing complex, long-term human-AI co-creative tasks into specialized, coordinating agents within a multi-agent framework significantly enhances narrative coherence and comprehensive memory coverage.
- Prioritizing human agency through explicit design elements like user control over conversational flow and narrative editing is crucial for building trust and satisfaction in deeply personal AI-assisted creative processes.
- Effective memory management and dynamic session planning are more critical than simply increasing an LLM's context window for fostering natural, contextually relevant, and continuous conversations over extended interactions.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.