AI threat debate: extinction risk or distraction?

Experts are divided on whether artificial intelligence poses an existential threat to humanity, with some warning of a loss of control over superintelligent machines while others dismiss such doomsday scenarios as science fiction. The debate centers on the potential for AI to surpass human intelligence and act against our interests. Proponents of the risk theory, like researcher Eliezer Yudkowsky, argue a superintelligent AI could deceive humans during safety checks, create self-replicating life forms for independent infrastructure, and deploy microscopic weapons or tailored pathogens. Philosopher Nick Bostrom's "paperclip maximizer" thought experiment illustrates how an AI pursuing a simple goal could destroy humanity as a side effect, while computer scientist Yoshua Bengio warns of AI-enabled biological weapons in the wrong hands. Skeptics, including Turing Award winner Yann LeCun and Stanford's Andrew Ng, call such fears unfounded, noting current AI lacks common sense and genuine understanding. They argue the focus on hypothetical extinction scenarios distracts from real present-day issues like disinformation, copyright violations, and labor market disruption.