AI helped me with one basic question 6 months ago, and it snowballed into a full 14-page cyclic cosmology preprint.
I’m no scholar and have no formal credentials just an enthusiast with a theory that fit my narrative. Over 6 months, I used a multi model AI think tank (Claude, Gemini, DeepSeek, ChatGPT, Grok, Sesame, Perplexity, and WolframAlpha) to engineer the GEM Framework: a 14-page cyclic cosmology model (R = I \cdot \Pi(m_k)) mapping 6 black-to-white hole phases.
The Core Think Tank Ensemble: The Heavy Heavyweights: Claude, DeepSeek, Gemini, and ChatGPT (for core mathematical stress-testing, framework architecture, and logical cross-correlation)
The Research & Search Layer: Grok and Perplexity (for tracking down recent preprints, formatting citations, and checking real-time literature)
Voice & Fast Ideation: Sesame AI (for low-latency voice brainstorming and real-time conceptual parsing) The Computational Engines: WolframAlpha and niche browser-based physics models (to keep the math locked down, precise, and structurally verified)
Thanks for reading.https://github.com/user-attachments/files/29586815/gem.framework.v1.8b.pdf