Progress (technological or otherwise) is exponential by nature, because it builds on itself. What has changed is that its pace has outstripped our capacity to integrate it. So our age of peak contribution trends lower, towards the younger generation, which has less to unlearn.

As I grow older, I can see how my thinking is affected by my past. Someone with an unencumbered perspective will do the new thing, when I cannot.

It's an economic problem. Right now, there's this torrent of new minds to come up with new ideas thanks to their lack of preconceptions, so there's no effort put towards alternatives.

There's a lot of possibilities that will open up as our tech advances. Say, wiping all memories of your field so you can learn it again with a fresh mind. Spawning short-lived digital clones of your mind with the mental 'temperature' turned up, in whatever way that's possible- maybe just dose them up with synthetic lsd in different doses- and observe the results. Lots of ways to get novel ideas without having to rely on wastefully growing billions of brains (and the attached bodies!) before throwing them, and most of the information in them, away to rot.

A specific technological change: writing (and printing.) We were able to remember longer spans more accurately than could be done by memorization and the passing of that knowledge from old to young. Reversions occur due to the active destruction of recorded history through war and religion. When the retention of novelty depends on the health of a relatively small number of people, as a society you're perpetually rediscovering instead of discovering.

Now our problem is even having the ability to assimilate enough established knowledge during a lifetime to have a chance of making any novel contribution. The obvious endgame to AI is a dream of being able to assimilate information and innovate indefinitely.