Hi, I am Vincent Liu, Chief AI Mathematician at WiseMotif.ai. This post primarily introduces our Grothendieckian Programme (a conceptual system and framework for Recursive Mathematical Discovery), focusing on three papers on our GitHub (an overview, 12 theses, and 57 open problems).
1. The problems stem from my seven-work series (seven works, 34 volumes). This series is quite extensive, comprising 26,021 pages that cover the architecture of reasoning systems as well as several research topics, and including 5 research topics about RH/GRH, BSD, Hodge, abc and Langlands Program, but none of the seven works claims a proof of the problems it studies. And the 57 problems also don't include RH/GRH, BSD, Hodge, abc and Langlands Program themselves.
2. I primarily designed the core conceptual framework, theoretical system, system architecture, and core algorithms, while AI assisted me with detailed writing and literature searches, ultimately enabling the completion of all seven works.
3. Since I have only recently open-sourced the entire body of work, it has not yet undergone peer review. I particularly welcome AI scientists, mathematicians, and engineers to identify problems that have actually already been solved, attempt to solve them, and supplement the formal statements in Lean, and we already have a Lean statement in problem 49, so that we can jointly build this new 'AI for Maths', thank you.
AlienMuse•30m ago
1. The problems stem from my seven-work series (seven works, 34 volumes). This series is quite extensive, comprising 26,021 pages that cover the architecture of reasoning systems as well as several research topics, and including 5 research topics about RH/GRH, BSD, Hodge, abc and Langlands Program, but none of the seven works claims a proof of the problems it studies. And the 57 problems also don't include RH/GRH, BSD, Hodge, abc and Langlands Program themselves.
2. I primarily designed the core conceptual framework, theoretical system, system architecture, and core algorithms, while AI assisted me with detailed writing and literature searches, ultimately enabling the completion of all seven works.
3. Since I have only recently open-sourced the entire body of work, it has not yet undergone peer review. I particularly welcome AI scientists, mathematicians, and engineers to identify problems that have actually already been solved, attempt to solve them, and supplement the formal statements in Lean, and we already have a Lean statement in problem 49, so that we can jointly build this new 'AI for Maths', thank you.