Recent searches
Search options
Well, last night's thread sure was (and continues to be) a wild ride.
I'll be blunt here: please believe me that I'm experiencing problems frequently enough to make me not like this place. And **please** try to understand how a repeated onslaught of suggestions on how I myself need to solve those problems doesn't give me much solace.
If I have to be responsible for managing this with whatever tools or strategies you throw at me just to be here happily, that's not something I need in my life.
Please feel free to tell me to pound sand. I don't need to be here, and "successful YouTubers" is hardly a marginalized group.
But also understand that I *want* to be here. I'm not leveraging this platform to spread my videos, I'm here to offer a way for people to connect with me. And I'm here to have fun.
At this point I need to self-censor a /lot/ to keep it fun. I'll try that for a while, but it's a damn shame I need to do it.
@TechConnectify I can’t imagine how much of a pain in the ass toxic people can be when you start getting even a little success on YouTube. You must feel like you wish you could hire someone to screen your comments.
@mos_8502 This is the thing though! This is the thing that I feel like nobody understands.
Despite the fact that I'm sure my YouTube comments are so much more toxic on the whole than what I get here on Mastodon, /I don't see them/
I have to go into my YouTube studio and look at comments in chronological order if I want to even notice that drivel. Otherwise, it gets filtered way down low and I never have to see it.
To be honest, that's what I want here. Some automatic ranking of quality
@mos_8502 but a lot of people are fundamentally against that here, admittedly for some good reasons.
On the other hand though, I don't think it should be thought of as a right to get your thoughts in front of somebody else. When someone is actually being problematic, enough signal makes it through the noise for that person to hear it.
Here though, everything is signal and nothing is noise. Which means it's all noise.
@TechConnectify For comparison’s sake, what was it like on Twitter for that sort of dickish commentary?
@mos_8502 I rarely ever saw it. It would be hidden under "more replies" which, to put that in Mastodon parlance, functioned like content warning.
And if it was bad enough, Twitter just wouldn't let me see it at all.
The people on the other side were mad that they were being shadowbanned or whatever, I'm sure. But it was a hell of a boon for my mental health and made the platform a lot easier to navigate than this one.
@TechConnectify @mos_8502 only in small bites (probably because I’ve retracted from here a bit myself) but I’ve experienced this too and immediately made this place way less desirable to be whenever a post gets “popular”. Fucking manifestos getting thrown out left and right that id never see over on bird app
@TechConnectify @EposVox @mos_8502 I'd love to know why people do this, do people not have anywhere else to pour their own unrelated frustration to and fixate on "hey this person has a platform lets go complain" or something?
@TechConnectify It’s a tricky problem. On the one hand, you want to hear what people have to say to you as a general rule, but on the other you want the abusive bullshit to get tossed in the garbage where it belongs. What most of the better instances do is they block a long, shared list of known toxic instances at the server level so nobody there can follow or interact. Then the admins of those better instances respond to reports of abuse by first blocking the abuser from the instance, and if that instance keeps being a problem, that instance gets blocked.
It’s not an ideal situation for the level of bullshit you’re dealing with.
@TechConnectify Very few instances are staffed to a level that would allow them to handle mass abuse coming from a myriad of otherwise unproblematic instances. One option (and it’s not a great option, nor is it an actual fix for the actual problem) is restricting your posts to followers only and requiring manual approval of new followers. That’s a lot more work than you should be asked to do.
Would blocking poorly-moderated instances be of any help?
I can't help but thinking of this as a technical problem (which I know isn't the full domain of what it IS, but that's how I think, problem-solving-wise), and the solution I can think of would be to have a client that actively filters content (or automatically blocks bad actors) and just doesn't show you the crappy stuff.
That way, people have no idea they're being ignored, and the way #Fedi works doesn't have to drastically change.
@rl_dane @mos_8502 that would work probably, but it feels like a blunt instrument.
Honestly, I think Mastodon needs to look at things like rankings and some sort of automatic detection of bad behavior like Twitter did. Maybe that's a much harder problem for this platform to tackle, I don't know, but I do worry that the "not built here" ethos is strong here.
@TechConnectify @rl_dane @mos_8502 firstly I want to say I love your work, and would greatly appreciate if you remain on Mastodon in one form or another.
I think the other issue is that few of the people working in the Mastodon universe, with the exception of some moderates and instance owners, have the same level of engagement that you might be seeing. Meaning they likely aren't seeing the same issues you are seeing.
The standard FOSS answer to this issue would be to submit a Pull-Request which fixes this for you, and others who have the same issue regarding scale. And this ignores the fact that not everyone has the knowledge, time or spoons to solve an issue they might be encountering.
I'll cut my toot short here to avoid being another manifesto in your mentions.
I do hope you'll stay.
@TechConnectify @mos_8502 this is exactly what I was thinking about last night. Today, the server side can drop toots per user matching specific words/phrases the user configures. I think this could be expanded into offering a low-tech (but performant) Bayesian filter too. This wouldn’t catch everything, but it would catch a lot of things depending on how intently you thumb up and down comments.
I don’t think this is controversial - it’s just like having a very long ignored words list.
@TechConnectify It might be possible to make this opt-in and local for your account. That is you enable the filter algorithm on your side by opting in.
Filtered posts will still there and shown to other people, but you don't get notified of them nor see them.
I don't know how Twitter handles this, but imagine it is similar.
@ryze I agree that it could be opt-in, but if I were to opt-in I would want other people to see replies in the way I want them filtered - because honestly that's a huge problem here.
There is no way for anyone to gauge quality of replies OR for people to signal quality of replies. They just stay where they are in order, and favorite count isn't even visible by default.
I understand the hesitation in making replies ranked etc. but to be honest for someone like me that's a necessity.
@TechConnectify I believe everyone should have an equal voice to the masses, introducing any signaling of "quality" of replies just makes it unfair. If you don't want to see something personally, that's fine, but it shouldn't affect others.
Even ordering by posting time isn't fair, because people who replied first get their post seen more often.
@ryze I agree with this sentiment, but it doesn't work at scale. This platform is unusable for me at the moment without some signal to noise filtering - and since everything is signal, it's all noise.
I think that has to be chewed on - if you are unwilling to implement filtering, that means people like me functionally cannot be here. And the behaviors that are pushing me away simply won't get any better. There has to be visible pushback.
@ryze And to be honest, the fact that you agree chronological ordering is also unfair tells me you understand this is a complex problem.
A decent filter will bring important signals through the noise. And truly, Twitter's filter was decent. I'm sure it sucked to be deemed noise, but right now the late replies will never be seen by anyone but the OP - and I'm begging you to understand that I desperately don't want to see everything thrown my way. It's not healthy.
@sxpert @TechConnectify AI is quite unstable at times and can get things wrong. Furthermore it will be biased depending on what dataset it was trained on.
AI is just statistics
You get it to learn what you consider spam yourself, just like with the spam filter in thunderbird
@sxpert @TechConnectify In case of adaptive AI, this will surely help a ton, but it should be implemented properly.
@TechConnectify Some amount of filtering will surely be needed at such scale, true. But we'll need to find a balance.
Perhaps allowing only a few replies to get "pinned", the amount of which could be configured locally?
As far as I understand right now, you want to see the most important and acceptable replies that is deemed by the majority first.
@ryze There's more nuance to it, but broadly yes.
I think people are largely so burned out by "characters of the day" and "getting ratioed" and whatnot that they're really hesitant to implement anything that even echoes those things.
But there's really no way for anybody to signal "this is a bad take" or "this is bad behavior" here. And muting/blocking, while it protects you the individual, does nothing to indicate right/wrong.
We need social cues on social media. There aren't any here.
@ryze So when designing a content filter, it doesn't necessarily even need to analyze the content in a post. It could be based on favorite (I'm not sure up/down voting a la reddit is a good idea but some sort of human inputs would help a lot).
This would also drastically cut down on the reply guy problem because the might actually see that their idea has been cosigned by other people /and they don't need to spew it out themselves/
@TechConnectify This will help true, but then this will be unfair for the people who got just unlucky and their posts weren't seen by enough.
Furthermore it can do more harm than good, since the right/wrong is dependent on the audience that decides. Reminds me of Flat-Earthers, which just get stuck in an echo chamber and barely see opposing views.
There has to be a a balance, I am not against introducing the rating system, but it shouldn't be the ultimate thing that is controlling how far your voice can reach.
Thinking of simple case, interleaving regular replies with replies pushed to the top by rating system might be a good balance.
@ryze It's nice to strike a good balance, but to be honest life just isn't fair. I have no doubt in my mind that my position is 95% luck - I caught a lucky break with my video on VCRs and that snowballed into a career.
Adding to that, people don't even agree on what fair is.
So... this might very well sound callous, but at some point something is going to get prioritized that leads to others getting left behind. I think it's good to work against that, but fixating on it leads to paralysis.
@ryze This isn't exactly related, but this is my whole thing about the FOSS movement.
Y'know why I use Adobe software? Mainly because when I encounter a problem, I will know other people have encountered the same and there will be resources to help me fix it.
It's not a perfect tool, and I don't particularly like Adobe. But actually defining a framework and making some rules we can all understand is often more important than quibbling over what is perfect.
@TechConnectify it's more of a problem of adoption and use of software more than it's freedom I believe. For example, Blender is FOSS and is loved and used by many and there are a ton of resources about it.
@ryze Sure, there are successful examples - but what they have in common is just that - commonality. People are using the same software and aside from little technical hiccups it's most the same framework.
Here, we are still deep within the "everyone wants/needs different things" stage and there's an obvious undercurrent of derision towards the more popular instances. There is so little direction outside of specific pockets, and imo nobody seems to *want* cohesion.
I’m going to chime in right here on this random thread because I don’t want my post to be too attached to the main thread.
I suppose you could tally me up as the person you described: a “hipster who doesn’t want anyone to find their fun coffee shop”
I’ll just say this. I agree with you. I don’t think Masto / Fedi is made for people with large audiences. I’ve heard complaints similar to yours from other people with larger followings.
Frankly, I hope it stays that way.
I used to work in corporate public relations where running big social media accounts was my job. I also used to run my own branded social media accounts with pretty sizable followings. About 4-5 years ago I left all of that behind for personal reasons I won’t get into.
Fedi remained though.
and it’s because of what you are describing that I stayed. I realized I honestly don’t want thousands of followers. I couldn’t keep up with them. If I had thousands of “friends” irl it would be overwhelming. To me, Fedi/Mastodon are a retrofit of real relationships that way.
Over the years, my relationships on fedi have migrated from one little pocket of friends to another. people and instances come and go. I really am just here to make friends. Fedi is, in my opinion, built in a way that allows that. Real human connections. The computers that run the fediverse can’t handle the fame that comes with big names.
I think the celebrities of the world would probably agree that scaling relationships to such a size is actually not that fun. There are tools in the world that can help us keep something of a handle on our personal lives amidst a high social status. But by in large, the scale of such an audience can be crippling. Both the positive and the negative sentiments that come out of it.
If you were as famous as Samuel Jackson irl then you would find it hard to go anywhere or do anything without being assaulted by intense expressions of love and hate everywhere that you go.
People weren’t meant to live like that imo. I’m still shocked that some choose it.
There are platforms built for this kind of sociability though.
Youtube, Tik Tok, Twitter, Facebook, Instagram– they all wan’t there users to feel like they need to be more popular. And they provide the tools and the features to help allow that. On these platforms you can rise in visibility, and hide all of the negativity that comes out of it.
Fedi won’t do that. In my heart of hearts I hope that it never does.
@TechConnectify @trunksapp has an option to kind of do this. You can sort comments by chronological and most liked.
"To be honest, that's what I want here. Some automatic ranking of quality"
An option to automatically mute people who are muted / blocked a lot.
@TechConnectify @mos_8502 Technically that should be possible on the Fediverse. Your instance could auto screen posts and be much more ruthless than other instances. I’m sure that this will be an option eventually because that anyone can just tag you in any post is untenable. It’s one reason why I have my own instance, to be in control (I pay someone for hosting and I understand it’s not for everyone).
@TechConnectify @mos_8502 It kind of sounds like you want an algorithm to keep the annoying comments away but for many here, we came to get away from the shitty algorithms that manipulate us. If that's what you want, go to the algorithm sites and talk to the sheep. If you value freedom of speech and not being manipulated, we would like you to stay here.
@timlocke Have you read the thread? I think your comment is a perfect example of what he’s talking about. Handling replies is easy for people like us that have a handful of followers. He’s got 34K followers and when he gets reblogged he ends up with 1,000s of knee-jerk replies mixed in with people trying to engage.
If you read any of his threads, he genuinely engages with his followers and replies. As if it were some sort of “social media.”
@markmevans I haven't read 1,000 replies. One of the great tings about Mastodon is the massive engagement which isn't found on other social media like YT. I can understand thousands of replies being overwhelming. From the threads I've read, there were no more than 100 replies.
To be honest, that's what I want here. Some automatic ranking of quality.This is something I've thought about on some ocassions but there is the fact that the decentralized nature of the fediverse makes this a bit tricky.
I very much want to see instances provide an algorithm for post ranking with the option to switch between using said algorithm or just displaying things in order.
The only main issue with this is the sheer fact that not every instance is going to have the same data to work with making it pretty much impossible to ensure everyone sees the same thing when looking at any given post. Its even worse on single user instances where they have essentially no data making any sort of algorithm close to useless.
