His recent statement that AGI may "only be a few thousand days away" is clearly an attempt to find a less quotable way of extending estimates while barely reducing hype. History generally shows that estimates of "more than five years away" are meaningless.