Martín Soto comments on The need for multi-agent experiments