The speaker says early arguments for interactive proof assistants based on checking correctness and details were not persuasive to the mathematics community.

“In the early days of interactive proof assistance, like I said, the argument that we need these because they're gonna check correctness and get the details right, that got no traction in the mathematics community.”