|
· Föderation EN Do 23.01.2025 17:38:23 @serge So, I'd appreciate from here on out if we keep our conversation public, since you're saying that I did not acknowledge your concerns or respond to your messages, which is unfair. I'm looking at our email thread, which has multiple multi-paragraph messages, and it seems to me that I've been responsive on the topic. |
|
Föderation EN Do 23.01.2025 17:46:59 In the public discourse you've yet to say "Yes this is a problem". You dance around it. You say it's early in the process, it's opt-in, etc. but I haven't seen you say a word about the fundamental issue, and the private dialog mirrors that. |
|
Föderation EN Do 23.01.2025 17:50:59 @serge There is no such thing as a single base level of truth, especially with social and political issues, and an ActivityPub extension that tried to establish that single base level of truth would be fundamentally flawed. Does that acknowledge your concern? |
|
Föderation EN Do 23.01.2025 17:56:42 |
|
Föderation EN Do 23.01.2025 18:04:02 @serge Yes, if an ActivityPub extension pretended to establish a base level of truth that applied to everyone, it would be harmful to marginalized people. |
|
Föderation EN Do 23.01.2025 18:05:04 That's all I've been looking for. "We need to be careful because this can cause harm" |
|
Föderation EN Do 23.01.2025 18:05:48 @serge Absolutely. |
|
Föderation EN Do 23.01.2025 18:12:52 You wanted thus public, so I've removed the user's name from the screenshot. These are the complaints about one of our users. In one post he said he thinks it's a terrible tragedy but doesn't think it rises to the level of genocide. But the reports are about almost everything they've said. Now see how a fact checking system could make things worse? What if it was run by the right and not the left, and we saw shit about lGBTQ+ people, or the kind of "Report your coworker for DEI" stuff that's going around the US government right now. I agree facts are important, but we must be thoughtful about what it means to a social network. Medien: 1 |
|
Föderation EN Do 23.01.2025 18:26:07 That is a frequent problem of the commercial platforms. Fact-checking is often very biased and don't take everything into account. It's frequently used to silence the marginalized and the political opponent. As we know, the UN are very biased against Israel. Far-right-groups use it to silence their opponents for calling their actions to what they are. Or greedy people, who live well by the current wellfare-system want to silence critics. They want to downplay the concerns of their (potential) clients and dependents. And so on. Fact-checking follows the money and the power too often. |
|
Föderation EN Do 23.01.2025 21:30:19 @serge No, I don't see how a fact-checking system would make all the reports worse. I am not sure you understand the functionality we're talking about. I did a couple of diagrams to help out. |
|
Föderation EN Do 23.01.2025 21:33:59 @serge Here's a rough wireframe of a current set of replies on a controversial post. The original poster's text is show first, and then the comments and replies are grouped together. It's a crude representation, but I want to highlight that a) people can comment and argue about the contents and b) fact checking by authoritative organizations is lost in the noise of other replies. Medien: 1 |
|
Föderation EN Do 23.01.2025 21:36:05 @serge Here's a common UI for fact-checking. It doesn't have to be the only UI available; it's just one that's commonly used. Fact checks are separated from the replies into their own section of the page, and additional information about how the conclusions were reached, and about the credibility of the fact checker are available. Also note that there's some curation by the receiver, their server, or their client, so one fact check is shown and the other isn't. Medien: 1 |
|
Föderation EN Do 23.01.2025 21:38:55 @serge The benefits here are that the review by a trusted entity is highlighted and shown right next to the original content. It's also separated from the comments or from quote posts. |
|
Föderation EN Do 23.01.2025 21:43:25 @serge So, one vector of harm is if the recipient uses an untrustworthy or biased fact checking service. The UI gives extra focus to the fact checking service, so it would elevate those biased opinions. |
|
Föderation EN Do 23.01.2025 21:44:50 @serge another potential vector, and I think this is what you're talking about, is that if a post gets this kind of fact check labelling, the original poster will be subject to increased pressure, including harassing comments and reports on their account. |
|
Föderation EN Do 23.01.2025 21:51:40 @serge For the first issue, I think the work of IFCN and others is helpful. Having qualified fact-checkers that follow industry norms and that are reviewed by peers gives much less of a chance of abusive and misleading fact check annotations. Of course, in a federated system there's no way to *make* people use those services instead of biased and untrustworthy ones, but I think it's possible to use the brands and organizations as a quick filter. |
|
Föderation EN Do 23.01.2025 21:53:55 @serge For the second, I'm not sure it's true. You're showing examples of reports made right now; I am not convinced that the volume will significantly increase with a fact-checking feature. It would be good to get some analysis from the field. I'll try and review it! |
|
Föderation EN Do 23.01.2025 17:48:25 @serge Again, nobody is trying to establish a ground truth -- one absolute truth that everyone agrees on. We are trying to find a way that established fact-checking services could provide those services across the Fediverse. There are many such services worldwide. https://en.wikipedia.org/wiki/List_of_fact-checking_websites |
|
Föderation EN Do 23.01.2025 17:55:39 Saying "Lots of services exist" doesn't acknowledge the danger. This kind of dogmatic refusal to acknowledge the very real concerns goes hand in hand with the proposal itself. Instead, you've shifted the topic, you've talked about how it's opt in (which is about as true as saying prayer in school i opt-in), and that it's early in the process. Nowhere I've seen have you said "I see how this could harm marginalized people." |
|
Föderation EN Do 23.01.2025 18:35:07 |
