SciMuse AI System Generates New Research Ideas by Analyzing 58 Million Articles
A study published in the journal PNAS on November 9, 2026, states that approximately one-quarter of the suggestions provided by the AI system named SciMuse, which analyzed 58 million articles, were met with high interest by experts.
The AI system named SciMuse, developed by Xuemei Gu and Mario Krenn, has generated new research ideas for scientists by scanning 58 million scientific articles. The study, published in the Proceedings of the National Academy of Sciences (PNAS) on November 9, 2026, highlights the potential of artificial intelligence to accelerate scientific discovery processes.
More than 100 group leaders from the Max Planck research community reviewed over 4,400 research proposals generated by SciMuse. As a result of the evaluations, 24.9% of the proposals were found to be highly interesting, receiving a score of four or five out of five. The average score for all proposals was recorded as 2.40 out of five.
Evaluation: The research results indicate that artificial intelligence can be used as an auxiliary tool in scientific idea generation, but human expert oversight is required to verify the quality of the generated content.
Sources
- haberton.com · Link