The promise and paradox of decentralization

But these don't actually need to exist. Decentralization doesn't have to mean a flat hierarchy, it could (against all trends) be a hierarchy with the user at the top, as social media once looked like. I could select the content I want to see by selecting the people that I trust and the content that they approve of, provisionally trusting the people that the people I trust also trust, and maybe extending that to an arbitrary number of hops. In that way I can choose my own mods.

This requires clients that work for the user rather than for the content creator, complete anathema to major browser vendors, who divide their time between anti-features to force their users to do things that they wouldn't want to do if given a choice, and candy for website owners/developers.

Labeling spam should punish the connections that caused it to be surfaced. That might damage serendipity slightly, but the modern web has run a bulldozer over serendipity with its engagement metrics algorithmically curating newsfeeds.

edit: one of the reasons I want (and probably many want) this kind of control is because I don't want spam, but I do want abuse. I don't want angry people filtered out as a rule, I want to be able to make the decision to blacklist specific angry people, or even people making stupid arguments in good faith. I want control over my own filter bubble. If that contracts my www to a gathering of friends and family, so be it.

If web 1.0 is simple content consumption, and web 2.0 is users providing content to each other filtered by powerful rent seekers, web 3.0 should be user governance. The "mod" system is an authoritarian lack of a governance process, and should be replaced with tools that aid collective decisionmaking.

It's solvable in the same sense that email spam is solvable. Much like adblockers, spam is "solved" by re-centralizing.

> If the volume of spam is truly extreme, then what's to prevent you from having distributed blocklists that are fed by automated processes (as opposed to manual additions), and users just subscribe to the ones that they trust?

Most users are unequipped to evaluate that and disinterested in putting in a bunch of work to defend themselves against the flaws of the system at hand. Like adblockers or email, most members of the general population want it to be easy, automatic, and require minimal effort beyond clicking the button that gets them going.

Users generally want things to work for them. Investing deeply in protecting themselves because the system's designers didn't consider abuse is rarely towards the top of the priority list. People like things that just work, and the further a thing is from that the more adoption will struggle.

Using Twitter as an example - they are going to be purposefully and continuously ejecting users. As long as the users are people who nobody wanted to listen to anyway then that is fine.

But the situation changes because sooner or later (potentially already happened) Twitter will eject actual community leaders and a community will need to reform somewhere else.

When that happens, open protocols and open platforms will help the community reform. It is quite hard to stop a community reforming on the internet for example because there are so many parallel communication channels that are basically free. So using Twitter isn't much of a risk because if they go bad then the community can reasonably move somewhere else.

People seem to expect "decentralised" leads to some sort of semi-utopia where everyone has equal status or act like some sort of saintly community driven by consensus. I don't see how it can work that way. It really just makes it easier to transfer power when the currently powerful start acting abusively.

That being said I don't think it's impossible to have such a system in a decentralized way, there are incentive structures you could build to handle this.

I skipped copying it all because it’s right there; not limiting my reply to that snippet.

Yeah information network signal attenuates over longer distances (space and time).

The internet created an ansible for assholes to hassle us.

Knowing this, fuck the open web. I want an email provider that just straight up blocks messages from senders I haven’t approved.

I’d like Signal to go the extra step and make communication over a 1:1 link, maybe using Wireguard, an option, as a fallback and to put other options in front of less savvy users.

Spam is just advertising by agents that won’t kowtow to government. I don’t need schemers in big biz or Africa.

Decentralizing of communication will never happen running some platform centralized around rugged individualist hackers motives. But we don’t exactly need FB, Twitter, etc, either.

Our phones are just TVs for business opportunities. I’ll just send my adventures straight to people who care and skip the noise thanks.