Skip Navigation

InitialsDiceBearhttps://github.com/dicebear/dicebearhttps://creativecommons.org/publicdomain/zero/1.0/„Initials” (https://github.com/dicebear/dicebear) by „DiceBear”, licensed under „CC0 1.0” (https://creativecommons.org/publicdomain/zero/1.0/)R
Posts
1
Comments
43
Joined
2 yr. ago

  • I follow some channels from lemmy and you see the videos as of they were posts.

  • This issue happens regardless of your way of dealing with it. If you have instances that don't allow AI content, how will they identify when someone breaks the rule?

    That's precisely why AI content is so dangerous in the first place.

  • Oh it didn't occur to me to check it on gitlab! I was afraid of being abandoned or something but it seems it's active and seems a good one. Hopefully more people discovers it since I think it deserves more attention!

  • Thanks for that reply! That was a great explanation :)

    I was worried the project was being abandoned or something but seems it's not. I'll definitely use that then! Hopefully more people discover it and gets more attention!

  • Fediverse @lemmy.world

    Personal website on the fediverse

  • That's awesome! I see some rough edges like some rendering issues in some models in examples but I will use this indeed!

  • Yes

  • I follow the linux experiment on Lemmy and it shows up fine too!

  • I didn't knew about Bridgy Fed but isn't that kind of creating a bot and duplicating profiles? I guess is better than nothing but it reminds me a bit of the app Surf which connects the fediverse with youtube and other stuff and while sounds good in practice I wonder if it makes people less willing to switch to the Fediverse because you don't need to anymore.

  • Aah I didn't knew this issue from Lemmy, really interesting. Creating a decentralized platform raises many new challenges that are hard to solve!

    Now I get why piefed approaches moderation in a different manner and tries to be more resource friendly.

    Thanks for the info! Really interesting stuff :)

  • Yeah one has to be careful with statistics and I couldn't find the whole Lemmy in one place so it's also not representative of the whole Lemmy.

    I thought ActivityPub did scale well and was ATProto (Bluesky) which had a lot more issues. I mean I can comment on Peertube using my Mastodon account meaning the whole Fediverse is properly connected and we are 1M MAU so I would say it already scaled good.

  • I don't know how to check for the whole lemmy but seems it's growing a bit but the MAU dropped a bit probably because of august: https://fedidb.com/servers/lemmy.world

    But the fediverse in general doesn't grow too much except when a scandal happens.

  • Yup indeed!

    I think talking openly about the issues to people and explaining there are better ways that exist is the way to go. If they are in with the idea entering the Fediverse is a no brainer.

    For me it's crazy how many public institutions are on private social medias and not open ones where they can set their own server and rules. It's also crazy how social norms have been decided by the top private companies in the world, if that's not a distopian I don't know what is it.

  • Great news!

  • There's no easy or hard way, if they are willing to join they will.

    Why are you in Lemmy? Explain your reasons to them and why you decided to join and hopefully they understand and join too, that's all you can do. Forcing people into the Fediverse won't make them use it of they don't believe in it.

  • If you follow a generic hashtag it quickly becomes too much, what I do is follow people and very niche hashtags that I know it won't bring that much content into the feed but that I'm interested in.

  • If you really don't care that much just go to Mastodon and leave bluesky as there is a better decentralized place already. If everyone on bluesky went to Mastodon instead the Fediverse would get a lot more attention and get better too.

    Specially seeing you in Lemmy, why did you leave reddit then (I'm assuming you did sorry if it's not the case)? For me was precisely to not give that much power to a single company and this applies to lots of things not just twitter. Centralizing power is bad and bluesky is exactly that at the end.

    I want to support people that really tries to solve this issues and not just people who takes this as an opportunity to be the next Elon Musk, ups I mean twitter.

  • First of all, as long as it doesn't replace search I'm fine with more options, I don't like when a company forces me stuff, that's it. I just think this is pointing to a future where search disappears and I don't quite like that.

    You might dislike a direct answer but young people don't. The design of the applications can encourage certain types of behaviors, can spread misinformation and promote racism as we have seen in many social media now.

    Things like Instagram and TikTok stress me out by the speed they show content and I could conclude no one will use those but then you see the new generations and how they even listen to music there, just the chorus of a song and go to the next one. This way of consuming information is being promoted by how the apps are designed; it is no coincidence that depression is increasing among young people.

    You can say that in principle people can use more responsibly this types of social media, but that's not the reality when we have algorithms trying to maximize the time we spend on the phone. Also, how do you know that companies don't add biases to LLMs? Turns out Gorg (the LLM on twitter) they have added to the prompt to not criticize Elon or Trump. That's the current problems of LLMs is that we don't have context and even if there is context, if you have to put effort into it people won't do it. This happens to scientific articles where people never check the sources, imagine with other stuff...

    I honestly like LLMs and I think they are fascinating and very useful in a lot of situations! And efforts like Perplexity gives me a bit more faith than google just throwing an LLM that suggests to eat rocks. And while you might see that eating rocks shouldn't be done, there's this bias that can be build behind any LLM that affects in a way that will be hard to avoid or notice. Same way current algorithms affects us more than we think they do and polarizes opinions.

    I mean will see where this goes, I just want companies to take the matter seriously.

  • I think it's simply that getting a direct answer is easier than reading different forums with different views and come up with your idea. That doesn't mean people want google search to stop searching. We have gemini, if I want to use gemini I can go to gemini. I don't get why everything has to be AI. We can have multiple tools, not everything out there is a nail.