wakest ⁂Slop coded fediverse servers are multiplying I guess we need to start tracking this somewhere!? What happens when there are more AI generated server frameworks then human written ones? The do-ocracy of implementation first protocol inconsistency sorta wins right now but that brakes down when there are more bots making the call then humans
Todd Sundsted@liaizon I think these are good concerns but I am concerned about the idea of “tracking”. right now my perception of the fediverse is that it has become by far the most overtly hostile social network I’ve been on with wrath regularly directed at “those who don’t conform” even if they diverge in trivial ways. and I am not the only one with a stake in this saying this! more of that made easier to direct is not going to benefit the fediverse in the long term.
@toddsundsted by tracking I mean a list on a website that publishes which pieces of fediverse software are not being designed by humans. Your worried about *this*?
@liaizon i perhaps don't understand the scope of what's being proposed. i assumed you meant a central website/repository for where these pieces of fediverse software would be listed, presumably not on on opt-in basis vs. something like the existing FEDIVERSE.md file that authors can choose to publish to self-track.
i have no problem at all with the latter. i think the former (the centralized site) has all sorts of problems, starting with the question of who gets to decide what goes on that list. opt-in is fine. opt-out (or no say in the matter at all) does not seem right to me for all the reasons that we don't like it when others do that to us!
nicole mikołajczyk@toddsundsted @liaizon I don't get the point against tracking them. Such a list would be useful even for 'vibe-coders' who want to experiment with fedi.
@mkljczk @toddsundsted exactly!
@liaizon based on your posts, you seem to be caring and thoughtful, but a boatload of the fediverse imo is not right now. just look at this thread. i desperately want mozilla to focus on building a better browser. but the thread devolved into a hate-fest, and that is 1) not going to bring mozilla around, 2) not going to bring the undecided over to the cause, 3) going to run off the newcomers who stumble into that, 4) turns my space here into a much more negative and unhealthy place. i predict that a list focused on tracking vibe-coded fediverse servers will quickly turn into targeting tool for the people i'm talking about here. we don't need that.
cc @mkljczk
@liaizon @mkljczk thinking about this more constructively, i believe very strongly that people should boycott products that don't align with their values, and i believe that organized boycotts can and have brought about needed change.
the problem i see here is that without a clear stated objective, the objective is going to become "harass them off the fediverse".
tallship@toddsundsted @liaizon @mkljczk
Honestly Todd, I'm not seeing anything more pernicious than one would have in their face just taking a casual stroll through Usenet or FidoNet - Truly; you could call them "curmudgeonville" and the comedy would spill until those vitriolic malcontents realize they're the butt of the joke... And by that, I mean these people too:
wrath regularly directed at “those who don’t conform” even if they diverge in trivial ways.
Those, lolz... wrathful ones.
I've seen so many well intended, ill-fated schema claiming to be closer to some imaginary ideal that satisfies the moderation needs for everyone, universally. Nope. nonesuch, I say. After decades administering listserv forums, y'know, the ezmlm's and majordomo's and Mailman mailing lists; or a coordinator for FidoNet, and participant in Usenet over UUCP even, all I can say is it will always, ultimately befall, as a responsibility the matter of moderation, to the individual participant themselves - as it has always been.
There's litterally no panacea, and this thread is evidence that the relatively nacent newcomer IFTAS really has nothing new to offer that hasn't been rehashed in one form or another.
The saddest thing I've noted has been the shift in web based forums since the demise of MySpace. Yes, really. Oh, for a while, sure you could post on Faceplant that you were sitting on your porch, picking your burgers from your nose and watching the grass grow and it would receive gufaws and 'likes' and the now retired +1 for those using Gplus; but eventually, there was a snide remark here or there and the descent into cacophony proliferated.
I hear people say these same things you noted, but about Bsky, or Threads, or you name it, there's that element of malignant nihilism that eventually becomes more pronounced the more it is fed.
People have more than enough tools, without abdicating their own responsibility to someone else's sensibilities, to moderate their own incoming mail, chat, forums, microblogs, commenting feedback systems. As if the actual spammers themselves aren't the most discouraging already. It's easy to mute or block miscreants, but for some reason people want to have someone else do it for them and capitalize on their own imagined frailty and feeble constitution.
There was a time back when USEnet was the main interchange of discourse along with Compuserve and The Well (Whole Earth 'Lectronic Link, among other contemporaries of the time) when it was simply understood by all that:
"The earth is not a safe place and there are no safe spaces"
That's never going to change.
Contrary to what those who refuse to accept responsibility for when it comes to going outside for a walk, bad actors abound, and they have always been everywhere throughout recorded history.
Some will claim that it's possible to turn over personal responsibility to some kind of bot or automated reputation model, and you yourself cite that people will dogpile other people who diverge, from sort of mob mentality in vogue at the moment.
An individualist eschews belonging to such ridiculous measures that would have one cede responsibility for their own welfare to nebulous machinations they don't themselves manage - when there's a very simple way in virtually every online system for the end user themselves to filter content stemming from personalities they don't wish to hear from twice.
Maybe I'm just wholly unimpressed with the feeble nature of modern enmity, because from where I'm sitting, it doesn't even begin to compare to the pernicious ad hominem of decades past.
Or maybe it's just become popular to play helpless and hapless victim and pretend to be gravely offended when some bad actor awards them the opportunity to claim offense? If so, and like every other fad, This too, shall pass.
@mkljczk @liaizon on the other hand, if the proposal is some kind of tool that allows the proliferation of new fediverse servers to better conform to the standards in place on the fediverse, then i definitely missed that! and it is a more interesting proposition, whether that proliferation is being driven by agents or humans.
@toddsundsted @mkljczk I have been keeping track of every piece of fediverse software that I have come across for many years. I help curate delightful.club and have helped with fediverse.party as well. A lot of people come to the fediverse looking for an alternative from the horrors of the status quo of social media. Many people pick their software cause it supports their ideological threshold of some sort.
@toddsundsted @mkljczk There are many many people here who would like to avoid using software written or designed by LLMs, and there is nuance here. There are people who are using LLMs to help them on with spesific features or integrated into their workflow, all the way to people who have never written software before now making fediverse tools. I believe people have a right to know how their tools are being built.
@toddsundsted @mkljczk There are those that will chose to try to stay "pure" and try to not interact with software built by LLMs and there will be people who want to play with the newest vibe coded things as soon as they get their hands on them. I am personally interested to see how this dynamic plays out. I am already seeing people getting mad they find out a tool they are already using is being built by an LLM and they can feel tricked. Like their trust in that thing has been taken away
@toddsundsted @mkljczk I have also played with vibe coding and I am not personally interested in shaming anyone for their use of new tools (even when those tools are being made or funded by literal fascists) but I think if servers or software are not being upfront with the fact that they are being written majorly by LLMs there is a valid reason to report that information so that people who otherwise wouldn't know can make up their own mind what to do with that information.
@liaizon @mkljczk i understand these perspectives. i literally deleted my social media accounts in january of 2017 and joined mastodon.social. i started building ktistec in march of 2020 because i felt like it was the only way i could ensure i had the experience i wanted. i deliberately picked the agpl-3.0 license for very ideological reasons.
so, yes, people should endeavor to use tools that align with their values. they should not, however, expect tool builders to comply with their wishes just because they wish it. and i don't see the upside in self-reporting AI usage for the average open source software volunteer.
@skyfaller suggested avoiding any software that does not explicitly state that it does not use AI. i think this is a much better way.
Marcus Rohrmoser 🌻Hi @toddsundsted @liaizon,
comparisons are a slippery slope and hostility is ugly wherever it shows.
I for one find inducing mental harm for profit and commercial censorships too hostile for me to be attractive. I see this in corporate media and it appears by design.
So what can we do? I recommend engaging in friendly communities and avoiding hostile ones. As IRL. And yes, tools have to improve, but don't expect technical silver bullets for social issues. (Authoritarinanism isn't either.)
comparisons are a slippery slope and hostility is ugly wherever it shows.
I for one find inducing mental harm for profit and commercial censorships too hostile for me to be attractive. I see this in corporate media and it appears by design.
So what can we do? I recommend engaging in friendly communities and avoiding hostile ones. As IRL. And yes, tools have to improve, but don't expect technical silver bullets for social issues. (Authoritarinanism isn't either.)
To #Flancia and beyond!@toddsundsted @liaizon FWIW I agree with this perception, there sure is a lot of ingroup/outgroup thinking and pressure to conform around here.
Emelia 👸🏻@liaizon i think this just tests the protocol and what people want to do with it to the limits, we're seeing vibe coded AT Protocol PDSes too!
Definitely highlights rough or under defined parts of specs & tests interoperability.
@thisismissem I generally find it interesting. Seeing that the people making these really have almost no idea of how federation works and are able to get posts federating using entirely prompt engineering is pretty incredible. But I have been noticing the people behind these projects have mostly been very quiet about the fact that they are not doing any of the coding themselves. And when people ask them how things work there's this awkward pause
@liaizon yeah, that's a bit of a trend with AI usage being something of a dirty secret. I'm ambivalent on it's usage, but personally don't use it for coding at present (sometimes I have to interact with it through github or Xcode though 🤷🏻♀️)
Personally it's the token cost and the fact I deeply enjoy understanding code as to why I don't use it.
@thisismissem I have tried some things and been impressed how far I have gotten but not had the courage to publicize any of the things I have done yet. I am not a programmer at all, have made websites my whole life and am fine reading some code to figure out whats going on but actually writing the logic I am pretty unable. The ethical issues of being constrained to using tools made by these horrible corporations is where I am mostly stuck on. I hope we see some real floss LLMs competing on code
@liaizon there are some good FLOSS LLMs like Haidra: https://haidra.net
@liaizon @thisismissem LLMs aside for a moment, i'm not sure i see this as a net negative. if you have a group of people who aren't programmers, whether because of where and when they were born, their genetics, their training, their economics, and make it possible for them to participate in the open source ecosystem, i think that's an amazing gift. maybe not on day one when they, like i did once upon a time, publish garbage. but eventually.
i just don't see how gatekeeping someone's ability to participate in life where and when they choose, especially in open source, is a good thing. are we really the people who should be telling them "no, you don't get to be part of this!"
Julian Fietkau@toddsundsted Is there not a parallel here to “why learn to draw if genAI can spit out pictures that are good enough,” or “why lift weights at the gym if a forklift can do 100× as much?” Results are great, but they don't have to be literally 100% of what we care about.
Lowering barriers to entry to programming is wonderful. LLMs have me considering caveats on ensuring that new approaches actually lead towards understanding and growth outside of the specific tool.
@toddsundsted As an example, I love Scratch to bits: https://scratch.mit.edu It's built and framed as an educational environment, but it's clearly capable, and its use demonstrably builds understanding of programming logic.
I believe that LLMs *can* be used in a way that leads to similar understanding and growth. I'm not sure if they are very good at it. As @liaizon notes, the approach easily tempts people into publishing things they have little to no understanding of.
@julian i will be the first person to say i don't have all the answers, but i literally started "programming" by typing the source code for games in 80's programming magazines into the home computers of the time so i could play the games. i definitely didn't understand anything i was typing at first—understanding wasn't my goal. but it led to where i am today. i had friends that did the same thing who did not go on to become computer programmers. one became a nurse. one became a graphic designer. none of us knew what we were doing but one of us went on and learned more about it. i'm pretty okay with that ratio.
@toddsundsted Heck yeah, that's how I started as well. 😄
Not everyone has to deeply know programming, but everyone is responsible for the open source code they publish. Back to @liaizon's point, there's always been shoddy open source code. But with or without LLMs, if you publish code that you can't answer people's questions on why it does what it does, that makes you a crappy open source programmer, and people would be correct to be wary about your software's reliability.
@julian yes! problems with crappy code and poor contributions predate LLMs by decades. my hope is that some number of people graduate from that to good developers using whatever tools kids use in the future.
@julian ok, scratch is pretty cool
@julian i think self-improvement for the sake of self-improvements itself is a wonderful goal. i think building software by hand for the sake of building software by hand fits in the category of self-improvement. it sure does for me, anyway! i hope we never end up in a future where no one does anything (cooking, painting, etc.) simply because there's no need to.
@toddsundsted What I was trying to get at with those examples is, do you call weightlifters gatekeepers for only counting your accomplishments if you lift the weights without a forklift?
Whether a vibe coder who can't read any programming languages is a “real programmer” is ultimately a question of semantics that you can answer either way, but communities defining themselves by their capabilities is not inherently gatekeeping. Some words describe proficiency or experience, and that can be okay.
Weightlifting is interesting because it's an area I know. Olympic weightlifting (and sports in general) very definitely have "gatekeepers" who set the boundaries or rules for participation. Olympics aside, weightlifting and bodybuilding both have had endless controversies about performance enhancing drugs all the way to using bench shirts and what counts as a "fair lift". Rules and gatekeepers are necessary to for fair competition. (And honestly, if someone wants to use a forklift to pick up a barbell and call themselves a weightlifter, weird but whatever... it's far from the dumbest thing people delude themselves into believing...)
I'm also fine with communities defining themselves in any way they choose, whether it's called gatekeeping or not. I think a sense of shared identity, values, and behaviors is inherent in almost any coherent community, and I think that's fine.
If someone never looks at the code an LLM generates for them, I personally don't think they're a "programmer" in any meaningful sense—and I think there are pragmatic reasons to worry about code quality if there's not a human in the loop.
But I would not argue that we should keep people who use LLMs from participating in or contributing to the fediverse merely because they use LLMs. And I think agreeing on fine-grained rules for inclusion or not is hopeless.
This gets to the core of what I'm trying to communicate overall. I think we should default to welcome people to fediverse instead of looking for reasons to keep them out.
@toddsundsted I also don't think LLM users should be prohibited from contributing to the fediverse. Notably, no one in this thread has advocated that position, so I'm not sure why we're talking about it.
Reading the thread from the top again, I think this exchange (although it has been polite and constructive!) isn't really going anywhere. You've said your piece, I have objected to the points I felt were objectionable, I reckon let's call it done. 🙂
Nelson@liaizon There has been a lot of pushback from LLM lovers against people pointing out who is producing slop.
I think the better approach is probably to list what projects have taken a clear stance against LLMs, because with the amount of money pushing LLMs into every piece of software, it's most efficient to assume that any software that doesn't have a policy against slop will become slop.
This is a good list of general software, with a few fedi servers included: https://noai.starlightnet.work/list.html
@skyfaller @liaizon i also find this better. it's opt-in but still makes it clear to anyone who cares how to best pick their tooling!
silverpill@liaizon I am already facing this problem in @weekinfediverse.
I think vibe-coded projects will be de-prioritized because most of them are not newsworthy and immediately abandoned by their authors. A project will be featured only if it survives and gets some traction.
@silverpill @weekinfediverse so do you add https://codeberg.org/sindum/fedibook to a wait list? and check back in next month to see if its actually still being worked on?
@liaizon @weekinfediverse I would simply ignore an obviously vibe coded project with no history.
Fedibook doesn't have a license, so it is not qualified for inclusion either way.
irelephantHow many are there
irelephantHow many are there
@irelephant I have about 5 or 6 on my personal list so far. (Counting ones that seem to have started from vibe first not already existing projects that have started using Claude)