Large language models surpass human experts in predicting neuroscience results

Interesting. I imagine bias plays a role in humans coming behind. Also I suppose having perfect recall access to all similar studies makes a difference