According to researchers at NYU, infants outperform current artificial intelligence algorithms. The neural-network models fail to capture infants’ knowledge.
The findings highlight fundamental differences between computation and human cognition, pointing to shortcomings in current machine learning.
The novel idea of putting infants and AI head-to-head on the same tasks is allowing researchers to better describe infants’ intuitive knowledge.
A human infant’s foundational knowledge is abstract and reflects our evolutionary inheritance, yet it can accommodate any culture in which those infant lives. The model developed suggests ways of integrating that knowledge into AI.
MORE