4 ms·

>What would be a network which allowed for more thoughtful discourse? Or at least the ability to more actively mute topics, threads, and groups of users while s

by Zhyl 7y ago

>What would be a network which allowed for more thoughtful discourse? Or at least the ability to more actively mute topics, threads, and groups of users while surfacing better content algorithmically? I would be interested in a parliament-like protocol for discussion, structured debate and reaching consensus. We currently have tools for proposing changes to text documents (such as pull requests) which could be applied to a community rules/laws or values system, but we don't have anything that can debate the changes, log objections, track resolutions and compromises in a structured way. I've seen places where community decision making is emergent (e.g. autobans when a users reputation drops below a certain threshold or votekick in games) but nothing that formalises the process, allows review of results, links decisions to overarching principles (or notes where a principle has not been followed due to circumstances).

jonshariat 7y ago
To add to this concept, it would be great if the elements of a quality argument[1] (i.g. Claims, Counterclaims, Reasons, Evidence, etc.) were built in. Also people could flag a comment (or parts of it) with a specific logical fallacy[2]. It would help guide people into arguing more fruitfully and better digesting what they read. 1. https://study.com/academy/lesson/parts-of-an-argument-claims-counterclaims-reasons-and-evidence.html https://study.com/academy/lesson/parts-of-an-argument-claims... 2. https://yourlogicalfallacyis.com/ https://yourlogicalfallacyis.com/
- pas 7y ago
  It would be great to show arguments as processes (data, model(s), evaluation/claims), and other arguments as specific problems with the process (logical problem, bias, low quality data, bad statistics, etc..) Ultimately everyone should be able to agree on the data, the model and the evaluation separately. The differences should boil down to differences in personal assumptions (different Bayesian apriori distributions, eg. different morals, worldview, thresholds for things, etc.) This should lead to questions about stuff that is either subjective and thus people can agree to disagree, or to simpler objective problems. (For example that data collection methodology leads to higher uncertainty than claimed, hence the whole argument becomes inconclusive.)
busymom0 7y ago
Won’t votekicks lead to echo chambers?
- cheschire 7y ago
  Every single system you come up with to stop trolls and other abusive behavior will enable echo chambers. Even upvote systems.
haecceity 7y ago
It's not always possible to reach consensus. Sometimes you just have to admit no one really knows and leave the problem to experimentation or further inquiry. Or maybe it's just undecidable. But most people don't have the humility you admit that.
pas 7y ago
What's missing is organizing and breaking down big questions in a structural way. And naturally quality controlling smaller parts should be easier. (Especially if QC itself becomes structured, eg. show someone a simple argument and ask whether it is biased or not; then meta check the bias moderators and update weights accordingly. This will probably show polarization, so maybe then that might allow to algorithmically pick sane centrist moderators.)

by Zhyl 7y ago

jonshariat 7y ago

To add to this concept, it would be great if the elements of a quality argument[1] (i.g. Claims, Counterclaims, Reasons, Evidence, etc.) were built in. Also people could flag a comment (or parts of it) with a specific logical fallacy[2]. It would help guide people into arguing more fruitfully and better digesting what they read. 1. https://study.com/academy/lesson/parts-of-an-argument-claims-counterclaims-reasons-and-evidence.html https://study.com/academy/lesson/parts-of-an-argument-claims... 2. https://yourlogicalfallacyis.com/ https://yourlogicalfallacyis.com/

pas 7y ago
It would be great to show arguments as processes (data, model(s), evaluation/claims), and other arguments as specific problems with the process (logical problem, bias, low quality data, bad statistics, etc..) Ultimately everyone should be able to agree on the data, the model and the evaluation separately. The differences should boil down to differences in personal assumptions (different Bayesian apriori distributions, eg. different morals, worldview, thresholds for things, etc.) This should lead to questions about stuff that is either subjective and thus people can agree to disagree, or to simpler objective problems. (For example that data collection methodology leads to higher uncertainty than claimed, hence the whole argument becomes inconclusive.)

busymom0 7y ago

Won’t votekicks lead to echo chambers?

cheschire 7y ago
Every single system you come up with to stop trolls and other abusive behavior will enable echo chambers. Even upvote systems.

haecceity 7y ago

It's not always possible to reach consensus. Sometimes you just have to admit no one really knows and leave the problem to experimentation or further inquiry. Or maybe it's just undecidable. But most people don't have the humility you admit that.

pas 7y ago

What's missing is organizing and breaking down big questions in a structural way. And naturally quality controlling smaller parts should be easier. (Especially if QC itself becomes structured, eg. show someone a simple argument and ask whether it is biased or not; then meta check the bias moderators and update weights accordingly. This will probably show polarization, so maybe then that might allow to algorithmically pick sane centrist moderators.)