Partly what everyone here has said already - the technology takes time to develop and starting from square 1 makes more sense than starting at the final iteration.

Although, on the other hand, I'd say part of it is "Planned Obsolesence."

Imagine you're a company that makes computer chips - you can instantly make the smallest most compact most efficient chip possible - and then what?
After everyone who wants a chip purchases one - what drives sales beyond the slow trickle of replacements and late adopters?

IF, however, you design a chip that will become obsolete in 2 years - because you have an entire multi-decade plan for how you will scale down your chips over time, you can sell all of those chips... then 2 years later, they're obsolete - your new chips are better... so you can sell all those chips to everyone who bought one before... rinse and repeat for a few decades and you've turned a one time profit into a long-term business model that will generate you billions.