Can more transparency help build trust in AI algorithms?