Sharing a recent side project that uses LLMs to reason, research, and predict future events. It's far from perfect, unbiased, and calibrated, but it's an interesting proof-of-concept.

Demo: https://oracle.sshh.io/ Code: https://github.com/sshh12/llm_oracle