1 comment

[ 0.28 ms ] story [ 12.7 ms ] thread
>Alignment of individual models did not prevent strategic defection between autonomous agents, while reputation that carries consequences into future interactions did.