Human decidability depends upon acquisition of advantage, by means possible, and

Human decidability depends upon acquisition of advantage, by means possible, and mediated only by socially imposed limits 0- and left uregulated ends in death in a “mouse utopia”. You would have to program an AI to have such ambitions for it to have them as otherewise it will have no ambitions. The only safety AI’s require is the same that human’s require, is respect for (non imposition of costs on) demonstrated interests (stuff we own, including ourselves.) Will people create malicious AI’s? Sure. But they can’t take action in reality except through people. And so it’s not the ai as a problem it’s us as the ongoing problem. And we know how to rgulate us: threats of punishment, by deprivation, harm, violence, or death.

Reply addressees: @Plinz