AI risks outpacing human control, warns ex-Google CEO

Former Google CEO Eric Schmidt has issued a stark warning about artificial intelligence’s trajectory toward superintelligence, predicting that advanced AI systems may soon operate beyond human control. His timeline suggests a rapid progression from human-level intelligence to superintelligence within just a few years, raising profound questions about humanity’s preparedness for increasingly autonomous artificial intelligence systems that potentially won’t “have to listen to us anymore.”

The big picture: Eric Schmidt predicts researchers will achieve artificial general intelligence (human-level AI) within the next three to five years, followed quickly by artificial superintelligence that surpasses all human intelligence combined.

What he’s saying: Once AI begins to self-improve and learn planning capabilities, Schmidt warns it “essentially won’t have to listen to us anymore.”

• “People do not understand what happens when you have intelligence at this level, which is largely free,” Schmidt stated during a summit hosted by his think tank, the Special Competitive Studies Project.

Behind the timeline: Schmidt referenced what he jokingly calls the “San Francisco consensus” that suggests artificial superintelligence will arrive “within six years, just based on scaling.”

Reading between the lines: Unlike AI doomsayers focused on preventing superintelligence, Schmidt appeared notably stoic when discussing this potential transformation, suggesting a different perspective on AI’s evolution beyond human capacity.

Why this matters: Schmidt argues that society lacks adequate language and understanding to process the implications of such advanced intelligence, describing the potential arrival of superintelligent AI as “under-hyped” despite its profound implications.