What's worst about social media comes not from the platform itself but the advertiser model. Algorithms are a way to drive engagement for advertisers. The longer you stay on the site, the more ads you see. The more ads you see, the more money the platform makes.
On a subscriber model, there's an entirely different incentive. They want you to be happy enough to keep your subscription, of course, but beyond that, they have no particular incentive to keep the load high on their servers.
Note, this is different than the problem in science publication. Scientists are not funded by readers, if we don't pay for a paper, science suffers. This is why the science publication market is such a corporate scam, especially where they're charging for access to research funded by taxpayers.
But unlike social media, newspapers could cure some of their woes, and ours, if they moved to a token system where, instead of subscribing to a particular newspaper, you subscribed to a reseller (at an upcharge) who gave you access to a group of sites.
Your annual subscription would buy you a number of monthly tokens, based on your plan, that could be exchanged at various rates for access to specific articles or the latest issue or the whole site for, say, 24 hours if you needed to do research.
Newspapers could set their rates. The Economist might want to charge two tokens for a standard article compared to the The Columbus Dispatch's one. They also might charge more for longer articles or for an investigative series.
They could also choose which resellers to supply content to, since there would likely be more than one, and could even sell to all of them. (Why wouldn't you?)
That means your subscription to any one reseller wouldn't give you access to EVERYTHING on the market, but it would give you access to a wider range of outlets than you might otherwise support individually, so smaller outlets have at least the opportunity to challenge the big onse.
More importantly, you would be paying for news, but you would only be paying for what you wanted.
There are drawbacks, to be sure. But a token system reintroduces a common market for information, which the world surely needs. And it is surely better than what we have now, where newspapers, following social media, are pursuing "audience capture," becoming more partisan (and less reliable) in an effort to drive high engagement.
Science does *not* suffer, did you mean to say? Elsevier does not fund science. Science does not suffer if the pimp is not paid for papers: see arXiv. Services like Elsevier used to be value added; now they are rent-seeking.
Papers written by researchers on publicly funded projects should not be paywalled. Again, see arXiv. Embargoed data from publicly funded research is a nuanced topic, because big projects often consume an entire career, and there have to be rewards sufficient.
Server run by the main developers of the project It is not focused on any particular niche interest - everyone is welcome as long as you follow our code of conduct!