Which experts exactly agree?

And on what would they agree, assuming they did (which they don't) since we don't even have a reasonable definition of Artificial General Intelligence (AGI) yet or an objective way to measure it.