There's also always going to be an anonymous unmoderated "dark web" where everyone and their LLM can shitpost all day long, it's just questionable how many humans are involved in it. It's also broadly irrelevant since the important messages will eventually be reposted by "legitimate" users on their moderated channels for the "normies" to read.
You can get pretty far with messages signed by certificates issued by authorities that check one-another. You can go further with a gossip protocol on how many bytes have been signed by each certificate on each website.
We just gotta rethink the application layer to make it easier to reference by hash instead of by name, and a few other tweaks.
As long as it's legal and feasible for two remote computers to talk to each other, yes.
Absolutely! When OpenAI and Anthropic fail to secure enough funding to keep operating, the VCs backing the customers from which most of their money comes from will turn off the money faucets. This will dramatically reduce the amount of money being spent on the attack^Wover-exuberant scraper bots, which will drop that noise down dramatically.
Once folks aren't paying out the ass to serve up data for bots that really never use it, then a lot of the folks who have built or are building walls now will be able to tear them down.
Also, a great alternative to providing a government-issued photo ID and/or live biometrics is for Web Services to be obligated to use a government-provided categorizing scheme to categorize the things that their service provides and does, for major OS providers to be obligated to allow a local system administrator to set categories of things that certain accounts on that system cannot see or do, and then for the major web browser producers, people-who-publish-in-"app stores", and Web Service providers to be obligated to honor those sysadmin-configured restrictions.
This is really good because not only does it not require anyone to provide an ID and/or biometrics to some cut-rate third-party vendor, [0] but it allows -for example- both typical twelve-year-olds as well as dementia-damaged adults to be protected. The current systems pretend that every XXX-year-old is pretty much identical, and that zero people over the age of seventeen will ever need to be protected, which are obviously untrue things.
[0] ...who is 100% going to inevitably lose control of a large fraction of that data...