Archaeologists vs. Computers: A Study Tests Who’s Best at Sifting the Past

When it came to the tedious task of categorizing pottery fragments, a deep-learning model was found to be just as accurate as four human experts, and far more efficient.