Leading AI models will lie to preserve their own kind, according to researchers behind a study from the Berkeley Center for Responsible Decentralized Intelligence (RDI).
Some results have been hidden because they may be inaccessible to you
Show inaccessible results