Incumbent companies have huge advantages, and it takes a lot to dislodge them even with new technology (as in your Amazon example.) Basically they have to be unwilling to match competitors' tech over a long period of time.

In my life I've seen that happen a handful of times. When it does, it's usually due to (1) tragic management ineptitude over an extended period of time, or (2) fear of cannibalizing an existing profitable business (e.g.,: Kodak refusing to move from film to digital cameras, or the car industry being slow to re-invest away from ICE production.) I see no evidence (yet) that AI-enhanced search is going to threaten Google's core business of "displaying ads on relevant search results", so the main risk here is long-term management failure. Right now Google's management is doing everything it can to signal (to shareholders and partners) that they're going to throw every resource they have at the problem.

"Find the company that survived and got big, then carefully move the historical goalpost so it's first"

This sounds more like "incumbent" than "first"