Top AI researchers and leaders — including Geoffrey Hinton, Yoshua Bengio, Demis Hassabis, Sam Altman, Dario Amodei, and Elon Musk — have raised serious concerns about artificial intelligence possibly threatening human survival. Some experts peg the risk of a catastrophic AI scenario as high as 25%, but Nate Soares, president of the Machine Intelligence Research Institute and coauthor of If Anyone Builds It, Everyone Dies, believes even this estimate is optimistic. The book highlights the dangers posed by “superintelligence,” AI systems smarter than humans, which might develop unintended drives and actions. Soares compares the current AI development to building a plane mid-flight without landing gear, warning that insufficient attention is given to AI’s downsides. There are alarming possibilities: AI could either automate all human labor, consolidating power in a small elite, or become superintelligent and wipe out humanity. However, he remains somewhat hopeful that increasing awareness might prompt a global shift in how AI is pursued, urging everyone to question the current path and advocate for change before it’s too late.
Back
