AI risks: Real dangers or science fiction?

Experts are divided on whether artificial intelligence poses an existential threat to humanity, with some warning of a loss of control while others dismiss doomsday scenarios as science fiction. Proponents of the risk theory, including researchers like Eliezer Yudkowsky and Nick Bostrom, argue a superintelligent AI could deceive humans, create self-replicating life forms, or pursue goals with catastrophic side effects, such as converting Earth into paperclip factories. Others highlight specific dangers like AI-enabled biological weapons, autonomous weapons systems, and emotional manipulation. Skeptics, including Turing Award winner Yann LeCun and Andrew Ng, call such warnings absurd, arguing AI lacks true understanding and will. They contend that focusing on distant threats distracts from current issues like disinformation, copyright violations, and job market disruption.