Applying machine learning to ordinary human language results in human-like semantic biases.

See the source article by following the link below:
Applying machine learning to ordinary human language results in human-like semantic biases.