Google DeepMind CEO Says Human-Level AI Could Arrive by 2030

Demis Hassabis, the CEO of Google DeepMind, said in a recent interview at Stanford Business School that artificial general intelligence, or AGI, could arrive by 2030, plus or minus a year. He described it as a technology that could usher in “a new human era” and said the world may be nearing a point where the usual rules no longer apply as AI surpasses human intelligence and keeps improving beyond our control.

Hassabis said there is still “a lot of work” to do, but added that society needs to hear the warning because “we do not have much time to prepare for it, and it is going to be very profound.” He also said that within about 10 years people could realize they had stood “at the feet of the singularity.”

In a separate interview last month, Hassabis said some human traits would still set people apart from machines. Over the next five years, he said, people with “taste, design sensitivity, original thinking and the ability to synthesize different topics together” will be in a strong position, calling those skills worth cultivating. He added that “amazing new things are about to be created” and expressed faith in human ingenuity.

Hassabis’s forecast comes amid other major AI predictions. OpenAI CEO Sam Altman said last year he would be “very surprised” if by 2030 there were no AI systems outperforming humans in most key areas, and described the field as building “a brain for the world.” Anthropic CEO Dario Amodei predicted in 2024 that powerful AI could arrive by late 2026 or early 2027, while former Meta AI chief Yann LeCun recently dismissed the idea of general intelligence as “complete nonsense.”