In the previous post I said that the view that writing is an old hybrid AI might have some bearing on whether robots take over, by which I meant, get better and better until they surpass us and become the main stream of civilization and history.
One optimistic view is that if AI keeps being designed as extensions to our minds, prostheses, implants or upgrades to our brains, then we, or "we," will stay in the driver's seat along whatever ride intelligence takes in the future. We could call this "takeunder." (Notice how "natural" the driver's seat analogy seems, even though a driver's seat is essentially a prosthetic fitting? Similar earlier analogies: in the saddle, at the reins.)
So, the idea that literature is a hybrid AI ("there, that wasn't so painful, was it?"), and the analogy between literature and Google+web, are examples of takeunder ... but what they prove about the future is hard to say.
I think writing is at least an example of how it could go well, or how it could be made to go well. Just think how we take art and writing-- information media external to our brains-- and call them "the humanities." That's a picture of what it could mean to be at home in cyberspace (& of course having a picture is reassuring). On the other hand, we're not guaranteed to have the sense, necessary understanding, foresight, style or grace to go that way.
Oh good, this gives me a seque into The Great Good Place.
Postscript: Above "I say," "We remember the things we've read as if [books were] a way to talk to those people again."
It's books that talked to, and were talked to by, those people. We call them our memories of our past, but it's books' memories, held for us. The authors talked to books, pretending to talk to an audience, and the books talk to us, pretending to be the authors. The arts of authors creating simulations of themselves, and of readers believing the simulations, are a symbiosis-- cybernesis?-- that has evolved and is evolving over time.