I'm not sure why you had that dig about ML skeptics but ML models can often perpetuate human biases if they aren't trained properly. For example, consider the amazon hiring algo that consistently rated female candidates below male candidates.