Musk warns superintelligent AI could persuade people, including himself

Elon Musk says on X that many people will be "one-shotted" by highly persuasive superintelligent AI, and that he thinks he might be one of them.

The post gives no examples, timeline, or mechanism for how such persuasion would work.

Separately, Rohan Paul says on X that plenty of engineers merge an AI-suggested fix because the model's explanation sounded right. The post offers no data or named study to support that claim.