Photo credit
InnoEval and new benchmarks show AI models struggle with original research
AI's limited originality in research could stifle innovation, urging a reevaluation of its role in fostering diverse scientific exploration.