Skip to main content


✅ Servers on Mastodon and the wider Fediverse use tried-and-tested methods to connect to each other, similar to the technologies on email and the world wide web. The network's servers are almost all run by non-profits supported by user donations. The network's software is free open source and designed for the needs of the people who use it.

🚫 Servers on Mastodon and the wider Fediverse DO NOT use AI, LLMs, blockchain, cryptocurrency, algorithms, trackers, ads or anything like that.

#FediTips

in reply to Fedi.Tips

> Servers on Mastodon and the wider Fediverse DO NOT use AI, LLMs, blockchain, cryptocurrency, algorithms, trackers, ads or anything like that.

That's a bit of a blanket statement, how could we know that's true?

in reply to C:\> LeFRACTAL ▊

@lefractal

AFAIK all the various software options Fedi servers run on are all open source, so outsiders can check what they do and how they do it.

Proprietary closed source software is pretty rare on here.

in reply to Fedi.Tips

But you can't be sure that the software runing has not been modified, or that some middleware is not doing something unusual with the data. I'm not saying it' would be bad. It's just that saying the Fediverse doesn't use AI or LLMS is just not verifiable.
Especially in "the wider Fediverse", as anyone can ActivityPub to anything!
in reply to C:\> LeFRACTAL ▊

@lefractal

Are you suggesting some additional steps people should take to verify?

Or are you saying that if something cannot be verified, it must be assumed to be untrue?

I'm not sure how practical that second option is in the real world?

in reply to Fedi.Tips

> Are you suggesting some additional steps people should take to verify?
Sadly it's not possible to verify it 100%

> Or are you saying that if something cannot be verified, it must be assumed to be untrue?
I'm not saying it should be considered false (untrue), but that something that cannot be demonstrated (verified) shouldn't be considered true.

The only right answer is "We don't know if some fediverse-enabled servers are doing things that others don't/bad things/use LLMS/etc"

in reply to C:\> LeFRACTAL ▊

If people are doing the right thing in how they run their social network instances, how do you suggest we give them credit for doing the right thing?

Or should we assume that they might be doing the wrong thing and give them no credit at all?

This entry was edited (1 week ago)
in reply to Fedi.Tips

You can give them credit by saying how cool it is that they spend resources running free software for all of us to enjoy,

You can give them credit by making a great account dedicated to helping users find their way on these platforms (that's you, kudos to you!)

But, saying things that we cannot know are true is not helping them, or others at all! It's just a feel-good fluffy statement that does not hold water, sorry. But, don't beat yourself up, it happens to all of us.

in reply to C:\> LeFRACTAL ▊

"You can give them credit by saying how cool it is that they spend resources running free software for all of us to enjoy"

How do you know it's free software? They might be running proprietary software? Why are you trusting them on that but not the other claims?

"saying things that we cannot know are true is not helping them"

If you were in a vegan shop, would you tell people they shouldn't bother because they have no way of verifying that the products are actually vegan?

This entry was edited (1 week ago)
in reply to Fedi.Tips

Well if someone tells me "Hey look at this Mastodon instance, I'm running it for everyone!" I'm gonna say "Hey great! Thanks!", not "Meh, it's possible you changed the code to track us all". Even, if, it *could* be the case, objectively.

It's completely different than saying "All fediverse instances everywhere are free of AI, LLMs and algorithms".

in reply to C:\> LeFRACTAL ▊

...but you don't know they are running Mastodon. It might be totally different software with a Mastodon-like front end (which some Fedi instances are, though usually another FOSS back end).

Why is it okay to take FOSS use on trust but not okay to take the other stuff on trust?

I'm sorry to be precise here but if you're going to criticise me for being imprecise, it seems important for you to stick to your own principles?

This entry was edited (1 week ago)
in reply to Fedi.Tips

We can come back to this later, I promise. But if you're okay with it let's focus on the other thread first so we don't spread ourselves out
in reply to Fedi.Tips

Ok, let's get back to this.

> "you don't know they are running Mastodon. I"
True. But here it wouldn't matter too much if they were lying, as they are not making any other claims about their "Mastodon" instance.

> "Why is it okay to take FOSS use on trust but not okay to take the other stuff on trust?"
Can you elaborate on what you mean by this?

> "it seems important for you to stick to your own principles?"
It's only important for me to stick to reality, really.
(continued)

in reply to C:\> LeFRACTAL ▊

Perhaps I've misunderstood, but you seemed to be saying that you would praise a site for running FOSS even if you couldn't verify it was actually running FOSS?

If it's okay to take the FOSS claim on trust, why can't we take the other claims (no tracking, no LLM, no blockchain etc) on trust too?

This entry was edited (1 week ago)
in reply to Fedi.Tips

The claim to be running free software is rather usual and innocuous. Because free software does not mean anything else than some software with a free licence. It's not claiming anything else about the software.

The claim that *none* of the services on the Fediverse use trackers, LLM, or blockchain is rather wild and precise. You're the one who made that claim, you should be able to back it up!

Just like "No Border Collie Bites" is a rather wild claim.

in reply to C:\> LeFRACTAL ▊

Let's say I have two Border Collie dogs. That breed is known to be rather friendly, and rarely bites.
But, it'd be wrong for me to say "Border Collies DO NOT bite", because it might get someone who trusts what I say to get bit in the future by some random Border Collie.
But, if I meet you one day, you have a Border Collie, and you tell me "You can pet him, he doesn't bite"', I will probably pet the dog. Even though it's *possible* you might be lying.
in reply to C:\> LeFRACTAL ▊

If someone says "water is healthy to drink and the best way to quench your thirst", there are all kinds of potential exceptions to that including fatal ones, but the statement is still accurate.

I don't want to take this thread into philosophy and linguistics, but my intention with the original post was something similar to a "water is healthy" statement.

Public messaging about important concepts tends to be along these lines. A lot of people simply ignore complex stuff.

This entry was edited (1 week ago)
in reply to Fedi.Tips

But what you said was more akin to "Any water you can find is good to drink!"

It's might be seen as a subtle difference, but it makes *a lot* of difference.

in reply to C:\> LeFRACTAL ▊

@lefractal

Public messaging is about getting important info to your target audience so that some net public good is achieved.

My account is mainly for non-techy new people who want help with the basics. My followers, especially the ones who ask for help, are almost all on Masto instances.

I'm trying to get across to these people that their server isn't tracking them or using blockchain or AI or anything like that.

You're possibly seeing my post from a different POV than its target audience?

in reply to Fedi.Tips

Kudos to you for the work you do with Feditips, I love to see people get involved in community work for free software.

Your heart is in the right place, and you're tuned into the intent and spirit of the Fediverse. But, that blinds you a little, to the point of posting misleading info. Simplification is not always a good thing.

...

in reply to Fedi.Tips

I don't think it's a question of point of view. The point of view of your audience is that they don't know much about the Fediverse, and will tend to trust what you say. <insert Spiderman quote about power and responsability here>
in reply to C:\> LeFRACTAL ▊

You're talking as if there is one effective message that is appropriate for everyone, but there isn't.

If you were promoting health in a country with good water treatment, a good message might be "Water is the healthiest thing to drink" because the main danger is unhealthy drinks with sugar or alcohol.

If you were promoting health in a country with poor water treatment, your message might be a warning instead about the dangers of water and their potentially deadly impurities.

This entry was edited (1 week ago)
in reply to Fedi.Tips

I'm really confused now, because I feel like you're now turning the argument around: you are the one who posted one all-encompassing message about all servers being safe.
The water treatment of the fedivese is quite good, but there might be bad puddles around, that look just like clean water. Hence why I was (still am) arguing for a more nuanced message.
in reply to C:\> LeFRACTAL ▊

@lefractal

I don't know what else to say except that I am trying to achieve a public good by explaining complex things in simple terms that are easily understood by someone who is new to this place.

I don't know of any Fedi platforms that use LLMs or blockchain etc. It's even less likely that a new person without Fedi experirence would use such a platform.

In theory there might be, but it's not helpful to assume that when it's so unlikely. In theory any water could kill you.

in reply to C:\> LeFRACTAL ▊

" I'm really confused now,"

I'm making the point that the biggest dangers aren't from some obscure Fedi platform that no one has heard of which might theoretically do tracking, LLMs etc.

The biggest dangers are that new people go back to some corporate platform like Facebook, Twitter/X etc that definitely does really bad stuff like tracking etc.

You tailor your message to take account of the most likely dangers your audience faces, rather than every theoretical danger.

This entry was edited (1 week ago)
in reply to Michelle Hughes

Well that's true but I have defederated Threads so no one on there is going to see this post.

Also, their connection to the Fedi seems pretty weak anyway as only a tiny proportion of their users connect to it and those that do are shut out from so many Fedi servers (especially since Zuckerberg announced hate posts are allowed on Threads).

This entry was edited (1 week ago)
in reply to Fedi.Tips

One thing I saw no one mentioning is that it's weird how the meaning of the word "algorithm" shifted from "a set of instructions to solve something" to "a non-open-source recommendation system made to make users addicted to a platform". In the original sense Mastodon has lots of algorithms, and in the newer sense it has recommendation algorithms (like the who to follow and the trending tab) but those are open-source and not meant to drive addiction.
in reply to Gustavo

Yeah, "algorithm" is a tricky word because it has very different meanings to different audiences.

I'm using the word in the sense that most people use it: an opaque set of hidden rules that hide stuff or push stuff or move stuff around usually without your knowledge or permission, a mysterious force.

Obviously this isn't the definition that compsci or developers would use, but the FediTips account is aimed at a more general non-technical audience.

This entry was edited (1 week ago)
in reply to Fedi.Tips

@Fedi.Tips @Gustavo

Your statement would have been more accurately stated nefarious propagation algorithms (I tend to think those words when a non-developer use the word out of context like this anyway, as I understand that they do not) though I am sure the non-developers that follow you will understand your context, while the developers will continue to cringe at the misuse of the word.

in reply to Fedi.Tips

to be fair I once tried to check the code for the who to follow algorithm: it wasn't easy to understand! I guess I could have looked for documentation instead, IDK...
in reply to Gustavo

@Gustavo @Fedi.Tips

Most non-developers really do not know what an algorithm is. They do not understand that there is no such thing as code without an algorithm, even if the algorithms are horrible (in the sense of inefficient, not nefarious). Sci-Fi books and movies commonly misuse the term. The word hacker was appropriated by Hollywood in much the same way.

There are processes that are non-algorithmic which do not break down into a series of well defined steps to achieve a goal (algorithm). Though coding is not one of them. To say a coding project has no algorithm would be a horrible insult to the developers. Though people have become used to only thinking of algorithms in the sense of Meta's Algorithm for propagating posts, etc. They have no clue at all that all code (at least professional/ expert level) code has many algorithms.

We need to get them all a copy of Algorithms + Data Structures = Programs by Nicklaus Wirth 😉

in reply to Fedi.Tips

IMO, FOSS and use-driven communities are the two pillars of the #fediverse, something you cannot find in commercial platforms. From those pillars flow everything else that makes the federated social media environmnet free, open and accountable.

Shut the #TechBros down I say. Migrate off their toxic platforms and let them die. I’ll take my chances with privacy (and my money if I choose to do so) on independent fediverse instances. #Resistance #antifa

in reply to Fedi.Tips

What do you mean by "do not use ... algoirithms"? Algorithms are a building block for all programming. Even how you display a chronological timeline is an algorithm. The trending functionality is an algorithm.

It's giving "our organic farming doesn't use chemicals" (however, objectively, water is a chemical).

in reply to Ira Peach

in reply to Unus Nemo

@unusnemo @ms_peach_says_acab

I'm well aware that "algorithm" in a technical sense means pretty much anything on the internet.

But my account is aimed at a wide non-technical audience, so I use language in a non-technical way.

I am using the word "algorithm" the way most non-technical people use it, to mean some opaque hidden code which moves stuff around, hides stuff you want to see, promotes stuff you haven't asked to see and generally messes around with your timeline without permission.

in reply to Fedi.Tips

@Fedi.Tips @Ira Peach

And I have acknowledged that in every comment I have made. Thanks for contributions to the Fedi, I truly appreciate it.

Though, you really cannot (or at least should not) expect developers to be happy with your propagating a useless meaning to a term. I understand why you used it. I am not offended. Keep doing a great job at what you do.

in reply to Fedi.Tips

@unusnemo

I don't think that's the way most people use "algorithm" when it's pertaining to a colloquial and social media usage. It's usually an uncountable usage, like "the algorithm" or "TikTok's algorithm", and not "algorithms" as you used it.

In other words, its usage is uncountable and metonymic.

Even ignoring that, there are better ways to word that specific usage for all audiences, like "(incomprehensible) curation algorithms".