There could be a hard ceiling to the capabilities of the current approaches that we hit at any time.

AI already does a lot now.

I am wary of anyone that says it's the most important invention ever. It could be "humanity's last invention" as they say (since future inventions would be made by AI), but it hasn't replaced huge swaths of knowledge workers yet that I've seen. So far, AI is a better version of StackOverflow and a proofreader and illustrator. These are huge things, but before we conclude that all lawyers will lose their jobs to AI, we should probably wait to see if even one lawyer loses their job. (Maybe one has, but you know what I mean.)

Also, I hesitate to buy into any innovation ranking. Consider: farming, writing, bricks, metallurgy, printing, transistor, internet, atom splitting--there are a lot of important inventions and they all affect one another.