The abuse-countermeasures argument is fairly persuasive, but possibly not entirely so. Not the least because scale also seems to greatly amplify that problems themselves.
Most forms of communications/informational abuse also seem to follow the propaganda / censorship / surveillance framework. Certainly spam, various forms of devoicing, and targeted harassment (bullying, brigading, doxxing, swatting).
Countermeasures fall into what I term hygiene factors, and much of my current thinking runs very strongly along epidemiological lines (and did so well before our recent viral symbiote emerged): addressing breeding reservoirs, transmission vectors, surveillance (of dysfunctional/toxic behaviours), susceptibility, treatment, isolation, and the like.
Distributed and decentralised approaches can work, and there are examples of these. From the world of spam: DNSBLs, SpamHaus, IronPort, pg's own Bayesian classifiers, and the like. How much added gain GD AI ML offers, and how much that relies of exclusive data monopolies, I'm not sure.
One further point that's going to be lost here, but what the hell: it's not entirely necessary of even possible to dismantle all informational (or other) monopolies. My key point has been to join the three ills which seem to be the subject of much present discussion as having a common enabling factor. And that discussing the issues separately or without consideration of etiology has little value. Sometimes the answer will be dismantling sometimes it will be to modify governance, ensuring public, social, and or commonweal orientation; sometimes to create counterveilling power blocs to balance forces.
Most forms of communications/informational abuse also seem to follow the propaganda / censorship / surveillance framework. Certainly spam, various forms of devoicing, and targeted harassment (bullying, brigading, doxxing, swatting).
Countermeasures fall into what I term hygiene factors, and much of my current thinking runs very strongly along epidemiological lines (and did so well before our recent viral symbiote emerged): addressing breeding reservoirs, transmission vectors, surveillance (of dysfunctional/toxic behaviours), susceptibility, treatment, isolation, and the like.
Distributed and decentralised approaches can work, and there are examples of these. From the world of spam: DNSBLs, SpamHaus, IronPort, pg's own Bayesian classifiers, and the like. How much added gain GD AI ML offers, and how much that relies of exclusive data monopolies, I'm not sure.
One further point that's going to be lost here, but what the hell: it's not entirely necessary of even possible to dismantle all informational (or other) monopolies. My key point has been to join the three ills which seem to be the subject of much present discussion as having a common enabling factor. And that discussing the issues separately or without consideration of etiology has little value. Sometimes the answer will be dismantling sometimes it will be to modify governance, ensuring public, social, and or commonweal orientation; sometimes to create counterveilling power blocs to balance forces.