3 ms·
I agree with the core thesis, and have had multiple discussions/debates about the importance of decentralization for promoting robustness through diversity (avo
by ssivark 6y ago
I agree with the core thesis, and have had multiple discussions/debates about the importance of decentralization for promoting robustness through diversity (avoiding monocultures).
A common complaint against federated structures (whenever I’m involved in this debate) is the claim that it’s much harder to fight spam & malicious actors than centralized systems operating at scale (which can afford to devote resources to this problem).
I find that claim quite unconvincing, and believe that the problem is manufactured by the existence of centralized systems. In decentralized systems, effects of malicious action (be it spam or misinformation of any kind) are localized and controlled — and cannot scale to have large impact (thus also providing some disincentives for malicious action).
- dredmorbius 6y agoThe 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.