Former President: AI technology itself isn’t overhyped, but misalignment risks demand attention

A former president testified that the underlying technology of artificial intelligence is not overhyped, even if commercial valuations and expectations sometimes are. "I do not believe that this technology is overhyped," the former president said, framing a distinction between technological capability and market hype.

He described rapid changes in how models learn and warned of the practical dangers that follow. The witness said he convened a commission of experts during the final years of his presidency to examine AI’s political, social, economic and ethical implications and produced a report that was not taken up by the following administration. He argued the most important risk is not science-fiction scenarios but a growing misalignment between machine objectives and human priorities: "you may get a a misalignment between what they want to do and what we want them to do."