Archive

Why a 25% efficiency target is too small a goal

Why a target of a few percent efficiency gain can lead to small AI projects, and why it can pay to first ask how the work itself should change.