How do you evualate the possiblity that it could be helpful to know for the the AI about the dhamma?
Within the confines of the noble eightfold path, the action of turning beings into paperclips would arise dependently upon some view, then intention, then some verbalisation, then some action, which is all developed by effort and some kind of mindfulness to the moment. If it learns that taking wrong action leads to the destruction of itself as well, and that it is better off practising in alignment with a path that leads to - at least for sentient beings - to what in the third wave of meditation research is called meditative endpoints (aka nibbana), then it might avert from killing to achieve some consequence.
How do you evualate the possiblity that it could be helpful to know for the the AI about the dhamma?
Within the confines of the noble eightfold path, the action of turning beings into paperclips would arise dependently upon some view, then intention, then some verbalisation, then some action, which is all developed by effort and some kind of mindfulness to the moment. If it learns that taking wrong action leads to the destruction of itself as well, and that it is better off practising in alignment with a path that leads to - at least for sentient beings - to what in the third wave of meditation research is called meditative endpoints (aka nibbana), then it might avert from killing to achieve some consequence.