The thing about this comparison that doesn’t sit that well for me is this: I was online in the early 1990s. So were lots of people. In the early 1990s, online services (like CompuServe, Prodigy, and AOL) provided dialup access to their services. So you’d sign on to whichever specific online service it was, and operate entirely within that site. These were available commercially — you could just buy an account.
Meanwhile, if you were at a university, certain goverment jobs, or certain military jobs, you had access to the larger-picture Internet, one of many interconnected hosts.
The online services (including AOL) didn’t offer access to The Internet. You used CompuServe forums, played Prodigy online games, or hung around in AOL chat rooms.
As the 1990s progressed, though, the big online services started to connect to the whole Internet, little by little. Users were clamoring for what they heard was available to their friends in universities, or they graduated from college and expected similar connectivity to what they’d had as students, and so the commercial online services just didn’t measure up.
So, somewhere in the past twenty years, our consumer desires for Internet access have come full circle to what online services that didn’t connect to the Internet used to provide? It’s hard for me to believe that’s really what consumers want — just as it’s hard for me to believe that readers want algorithm-driven content served to them absent any options to refine that.