OpenAI’s feud with mathematicians is only escalating
Twenty-five leading mathematicians signed an open letter arguing that AI labs are threatening their intellectual work.
A group of 25 prominent mathematicians has penned an open letter warning that artificial intelligence companies are jeopardizing their scholarly work by racing to outperform each other in solving complex mathematical problems. All 25 signatories have been honored with the prestigious Fields Medal, the top award in the field of mathematics.
This week, NYU professor Tristan Buckmaster alleged that OpenAI exerted pressure to withhold recognition for a collaborator who worked for Anthropic in solving a significant mathematical puzzle, questioning whether the company utilized the Anthropic researcher's work with OpenAI's Codex to generate its own groundbreaking proof over an intense weekend of computations.
OpenAI subsequently rescinded its sponsorship of a mathematical event at CalTech following criticism from university researchers. The authors of the new letter contend that while AI models potentially offering solutions to significant mathematical challenges could be beneficial for humanity, they will only do so if these discoveries are comprehensible and communicable to the mathematical community and wider society.
They pointed out that many such findings are prematurely disclosed, precluding adequate explanation, isolation of novel methods and ideas, and proper citation of previous work. OpenAI's unverified proof exemplifies this issue, raising concerns about attribution and plagiarism in creative endeavors.
Mathematicians worry that the absence of these collaborators in the development, integration, and eventual integration of AI-conceived ideas into the mathematical canon could stifle the vital human transmission of knowledge. As AI frontiers labs identify potential discoveries, they may spend millions of dollars utilizing large language models to beat original researchers to a proof, creating a culture of secrecy.
This letter comes in the wake of the Leiden Declaration, which also addressed the implications of AI-generated proofs on the mathematical community and offered recommendations for various stakeholders. Just as AI tools reshape workflows in fields like software engineering, mathematicians seek to highlight the importance of the intellectual framework that supports students, generates new questions, and connects these advancements to broader human civilization.
The letter warns that the challenges faced by mathematicians today echo issues faced by other scientific and creative fields, emphasizing the need to ensure AI advancements do not undermine the core purpose of human work.
Written by urgent.news from TechCrunch's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
