According to TechFlow, Andrew Kang, co-founder of Mechanism Capital, said that the combination of AI and encryption technology brings revolutionary opportunities to scientific research paper review. The encryption project @yesnoerror is developing an AI model to review 90 million scientific research papers. So far, more than 1,700 papers have been reviewed, and the error rate is about 3-4%.
AI can complete the literature processing work that requires 45,000 man-years in a few weeks, at a cost of only 1% of traditional manual auditing, about $30 million. The project is developing a model to assess the quality of papers and generate a standardized quality score for each paper, taking into account factors such as methodology, logic, and data integrity.
This will help distinguish high-quality papers and may promote better scientific research by establishing a ranking of universities and research institutions. In addition, the AI model is expected to revolutionize the peer review process and may eventually replace manual review. Kang said that he has been supporting the project operation behind the scenes and his position is not the focus.