what do you think about this feature?
What do you think about this?
cross-posted from: https://lemmy.world/post/25639684
The tesseract Lemmy app, has a little overview from mediabiasfactcheck.com (MBFC). It seems like a clever way to foster a healthy community.
If you click on the ranking you get details.
Its problematic to put the political center as least biased and objective.
Isn't this that annoyingly biased bot that was around months ago?
The bot used the same site for its ratings, so yes.
Seems like it
Yeah I saw that thread. It looks extremely controversial so I will not be implementing this feature.
The idea isn't bad, but the data source is almost comically skewed, same as for the Lemmy bot.
What is the baseline for this thing that it's labeling CNN "left"? It's practically the dictionary definition of centre-right.
No idea
I hate these bots and instantly block them whenever I see them.
It has no value to me. Deciding the line in the sand is just as much of a power play as the onerous establishment. Calling the US Democrats Left is disingenuous. They are on the right by a considerable margin. The USA has no Left. So this is only pressing the establishment and validating it.
If you want to potentially sidestep some of people's frustrations you might consider just using the credibility rating and focusing on whether a group provides factual reporting, rather than left or right of center
You can also create a user experience that more carefully manages expectations of the feature by having it be opt in, but presenting the option to users when it becomes available. That gives you the opportunity to give a short blurb acknowledging its imperfections and also highlighting its potential value
As someone fairly to the left wing myself, the fact that lemmy is so left wing is both a blessing and a curse. I don't see Nazis around, but being in an echo chamber isn't great for your ability to engage with perspectives other than your own, and makes you succeptible to narratives that reinforce your existing views regardless of whether they're accurate
I'd love this feature, in spite of its flaws, but it does definitely have them. Its based on the US overton window, which will frustrate folks from other parts of the world who may already feel lemmy sometimes forgets the world beyond the US exists. And the US overton window is changing as a product of the trump administration which may warp mbfc results, which could honestly be really dangerous.
Focussing on the factuality and credibility might help you sidestep those problems and make a feature people would find less frustrating, potentially even to the point that you could make it opt out.
Generally I appreciate efforts to build healthier, less echo chambery discourse, thanks for the work you're doing ❤️
Hmm I double checked how trustworthy their factualness ratings are and it seems ok and generally not controversial.
On the other hand it looks like they are insanely restrictive about their API, requesting a $10/month fee and only giving a API limit of 50 requests/month. There are a few scrapers that I've found but it looks like if I wanted to add this feature realistically I would need to (1) use a scraper to compile all their data into a database (2) host the database myself and (3) create my own API against this database. This is a lot of maintenance work, the biggest by far is maintaining the scraper. It's also a bit concerning how legit this would be since I'd imagine they would not be happy if they found out I was essentially scraping all their data and serving it myself.
Due to all of this I don't think I will add this feature unless I found a better way to access their data that wasn't as sketchy.
###Addition
Apparently the bot I showed isn't the best one. I was more talking about the idea behind it instead of using the same bot. We can find alternatives. Recommendations are welcome
I think a community notes feature would be better. Because community notes target specific statements in the article/comment, whereas this just provides a vague and hard-to-interpret score.
Along with others in this thread, I don't think this feature would foster a healthy community - it would foster an echo chamber. I would rather see an analysis of the language of the article pointing out any logical fallacies used, weasel words, etc. than a "left-o-meter". I have my own one of those based on my actual beliefs, not what someone decides my beliefs have to be to fit into the blue box.
So it's a meter based on how left it is? I guess I didn't take a good look at it, I thought it was something like something that checks the credibility of the website.
I mean, it's a little more than that, but only a little. There's a literal meter in your screenshot of how left or right the source is. "Credibility rating" requires us to trust MediaBiasFactCheck's credibility rating system, which I don't know enough about at the moment and so default to not trusting it.
That's why I say calling out the problematic structures would be more helpful - people could see it for themselves, right there in the article text, and then maybe also identify then without help later. This would foster healthier discussion than an echo chamber where people ignore a source based on its biases.
Of course some sources would be more note than content, but then some sources have argued in court they're not really news.
To answer your question in relation to Summit, or any Lemmy client in general:
Implementing a feature like this interferes with user agency. I dislike the idea that my device would be sending my reading habits to a fact checking site. Since the implementation would most likely be domain agnostic, that website wouldn't just know my political stance, it would also know my interests. I don't need any more targeted advertising for Legos, sports paraphernalia, or enterprise server equipment.
From a developer perspective, I bet that API costs money eventually. It's a trap.
lemmy.world admins made a bot that used MBFC and many considered it a bad move. I'm sure there are others but here is a thread on the topic. https://lemmy.world/post/18073070
What do I think about something telling me preemptively how to feel about a news source? That's not fact-checking, I'd rather have the news COMPLETELY stripped of source. This paternalistic bullshit can fuck right off, it's always going to be reflecting someone's agenda. Since you asked.