Key points are not available for this paper at this time.
The paper discusses the prospects and risks associated with the development superintelligent artificial intelligence (AI) and artificial consciousness (AC) ten years after Nick Bostrom explored these ideas in his influential book (Bostrom, 2014). We argue that the debate about the evolution of AI has changed at least in the following aspects: First, the development of superintelligent machines is no longer limited to speculations about a distant future – recent advances in this field already produce immediate and palpable impact across various sectors, including education, business, and technology. Second, we argue that the debate on whether AI systems might one day achieve a form of artificial consciousness (AC) has shifted from a theoretical possibility to a pressing concern. After reviewing some of the most likely prerequisites for the development of potentially dangerous AI systems, we suggest several directions of action to avoid the risk of losing control over superintelligent AI.
Șușnea et al. (Fri,) studied this question.