A recent study published in *Nature* found that large language models can predict the results of social science experiments. The article references multiple studies on the capabilities and limitations of AI in social science research, including work by Bail (2024) on whether generative AI can improve social science, Grossmann et al. (2023) on AI and the transformation of social science research, and Luo et al. (2025) showing that large language models surpass human experts in predicting neuroscience results. Other cited works examine whether large language models should replace human participants (Crockett & Messeri, 2024), the perils and opportunities of using them in psychological research (Abdurahman et al., 2024), and their use in simulating human samples (Argyle et al., 2023) and synthetic replacements for human survey data (Bisbee et al., 2024). Additional references cover topics such as which humans are simulated (Atari et al., 2023), whether AI language models can replace human participants (Dillion et al., 2023), large language models as simulated economic agents (Filippas et al., 2024), using cognitive psychology to understand GPT-3 (Binz & Schulz, 2023), LLM-generated messages persuading humans on policy issues (Bai et al., 2025), the levers of political persuasion with conversational AI (Hackenburg et al., 2025), human-level play in Diplomacy (