Lvxferre

Alright. Sounds fair. Instead of taking dangerous actions, I’ll make it create a report instead.

Thank you! Frankly, if done this way I’d be excited to use it ASAP.

Automod is just a tool, indeed, but how a tool is designed dictates or at least encourages its usage.

Trying to automate things and decrease mod burden is great, so I don’t oppose OP’s idea on general grounds. My issues are with two specific points:

• Punish content authors or take action on content via word blacklist/regex
• Ban members of communities by their usernames/bios via word blacklist or regex

1. Automated systems don’t understand what people say within a context. As such, it’s unjust and abusive to use them to punish people based on what they say.
2. This sort of automated system is extra easy to circumvent for malicious actors, specially since they need to be tuned in a way that lowers the amount of false positives (unjust bans) and this leads to a higher amount of false negatives (crap going past the radar).
3. Something that I’ve seen over and over in Reddit, that mods here will likely do in a similar way, is to shift the blame to automod. “NOOOO, I’m not unjust. I didn’t ban you incorrectly! It was automod lol lmao”

Instead of those two I think that a better use of regex would be an automated reporting system, bringing potentially problematic users/pieces of content to the attention of human mods.

That’s a good text. I’ve been comparing those “LLM smurt!” crowds with Christian evangelists, due to their common usage of fallacies like inversion of burden of proof, changing goalposts, straw man, etc.

However it seems that people who believe in psychics might be a more accurate comparison.

That said LLMs are great tools to retrieve info when you aren’t too concerned about accuracy, or when you can check the accuracy yourself. For example the ChatGPT output of prompts like

• “Give me a few [language] words that can be used to translate the [language] word [word]”
• “[Decline|Conjugate] the [language] word [word]”
• “Spell-proof the following sentence: [sentence]”

is really good. I’m still concerned about the sheer inefficiency of the process though, energy-wise.

Home pages? Can we please have the ⟨blink⟩ tag and the “UNDER CONSTRUCTION” .gif again? Those were of utmost importance!

Okay, serious now. I might be wrong but I think that the whole internet is going full circle, and that what the link describes towards homepages is part of a bigger process, of re-decentralisation. It isn’t just about getting news from homepages instead of social media; it’s also about how we find content (again, through human recommendation) and who owns it (individuals or small groups, as the ad “industry” is going kaput). It don’t think that’ll be exactly the same as the 90s/00s internet but similar in spirit.

Removed by mod

You’re exemplifying OP’s #1 and #2: a black person, who’s in a position to know what people will/won’t take offence with (since she’s a mod), is using the terms whitelist/blacklist, and telling you “it’s fine”.

But you’re still trying to argue it on hypothetical grounds? Shut up and listen to her dammit.

Choosing weaker forms sounds sensible - my criticism is which ones.

Many people react that way but think about it a little more. It’s a fact. Mutliuple Black people have proven it repeatedly.

Yup, I know that it’s a fact. You aren’t being fallacious, but the way that you phrased it sounds like that fallacy, so it’s a matter of clarity.

It’s the same deal as the “post less”, you know? People are misunderstanding you.

[from the other comment] That’s a great point, can I quote you on having seen it on Lemmy quite a few times?

Feel free to do so! However keep in mind that I didn’t really keep track of them, so if someone says “do you have proofs?” I have no way to back it up.

In the context, the author isn’t saying “you should reduce your whole Fediverse activity”. It’s more like “when talking about this stuff, if you aren’t black, think before you say something. And you probably don’t need to say it, it’s better to shut up”.

It’s sensible advice even if worded poorly.

[Replying to myself to avoid editing the above]

Another point that I’d like to highlight is that a lot of the racism in English is proxied through linguistic prejudice, due to the existence of racial varieties like the African-American Vernacular English ones. For example, picking on people who use habitual-be, or specific words/expressions common among AAVE speakers. It is racist and I’ve seen it here [in Lemmy] quite a few times.

What a wonderful user experience! No rice is fine if connection is down, right?

I like where this text is going. As such, my criticism here is mostly on better ways to convey some points, plus additional info.

Title + Intro

Subbing “start making” for “make” highlights better that every little change matters, and is easier to read.

In this context “more welcoming” says the same as “less toxic”, but the former should be better to “sell” to the readers the idea that they can and should contribute. (Plus the word “toxic” is bound to make some people roll eyes and ignore the message.)

1. Listen more to more Black people

It would be great if your text addressed people who shut up marginalised groups while claiming to speak in their names; it sounds a lot like “I’m an ally so chrust me, you don’t need to listen to [group], lissen to ME! ME! ME! instead.” I’ve seen this too often in social media, including here. Black people probably have a lot more to talk about this than I do.

2. Post less – and think before you post

Simply saying “post less” is bound to rub off people the wrong way, specially when removed from context (plenty people won’t read the section past that), as plenty people are aware that the Fediverse needs more content.

Sadly I’m not certain on a good way to rephrase this without erasing the message. (Perhaps merge it with #1? Just an idea.)

Stop asking Black people for evidence [… whole paragraph]

I believe that the conclusion within this bullet point is accurate and moral, but the whole package needs some serious rewording.

IMO a better approach here is to highlight that all those “excuuuuse me, where are the proofs that you’re subjected to racism in the Fediverse?” are a form of sealioning, regardless of the intention of the people asking it. Black person be asked once, they provide the bloody proof; be asked twice, they roll their eyes but still do it; be asked for the 1000th time, they get pissed and leave.

I’m saying this because, the way that this point is currently worded, it sounds fallacious (inversion of the burden of the proof). And even if most people can’t quite identify fallacies, it still ticks a lot of them off, they know that “something” is wrong.

Stop telling Black people that they’ll experience less racism if they change instances […]

It’s actually worse: it’s a form of racial segregation. It’s like telling them “you won’t experience racism if you sit in the back of the bus”.

Black people should feel comfortable to use the same spaces as everyone else.

Stop saying the fediverse […]

I think that this bullet point is perfect as it is. Just commenting on the underlying issue:

A lot of people here confuse personal experiences with general statements. Even if the Fediverse, in general, was friendlier or nicer towards marginalised groups, it doesn’t really matter when someone is pissed and trying to vent their bad experience, you know?

Also ablism

Just highlighting a typo. No issue with the message.

I’m afraid of both, too, but if I had to choose:

• evil but smart people only harm you when it benefits them
• dumb but good people harm you all the time

As such I see the later as far more dangerous. Although there’s an even worse group, that Altman likely belongs to - the dumb but evil ones.

Pretty much. But instead of adjusting it like “cook it for less/more time”, you say “it’s raw/mushy”. Or at least that what I think, based on the product info, but I might be wrong.

And… yeah, it’s all pretend. Just like “smart” some years ago.

• Advanced fuzzy logic technology with AI (Artificial Intelligence) “learns” and adjusts the cooking cycle to get perfect results
• Superior induction heating (IH) technology generates high heat and makes fine heat adjustments resulting in fluffier rice
• “My Rice (49 Ways)” menu setting – Just input how the rice turned out, the rice cooker will make small changes to the cooking flow until it gets to the way you like it

Based on the description the so-called “AI” simply adjusts time based on user feedback. That would be hilarious if not so sad as a marketing device.

Not when it comes to the bans. They’re also issued on a whim but there’s that implicit knowledge that you’re going to reset your IP, delete your cookies, and the ban will last exactly five minutes. Even if it’s supposed to last six years (like the time that I was banned for asking “I’m 12 and what is this?”, and the newbie mod wasn’t aware of the meme).

(USER WAS BANNED FOR THIS POST)

*slow clap*

I’m glad that I never even touched Something Awful, I went straight for 4chan. A platform that allows moderators to disguise an effective permaban (100kh) as if it was a tempban (“probation”) deserves absolutely no respect.

Finally, Etherwind was right. Wraith’s 20th Anniversary Edition didn’t come out in 2016.

AND THEY HATED HIM BECAUSE HE TOLD THEM THE TRUTH.

I find interesting that you’re using the bot for something else (coding aid) than me (translation aid), and yet the conclusion is the same - you need previous knowledge of what you’re asking to sort out the bullshit being generated by a system that doesn’t really know anything.

You’re describing the “plateau of productivity”, which is built into this theory:

Not quite - the plateau represents all useful ways that the tech can be used; what I’m describing is not even a fraction of that, it’s only the stuff that is too obvious to pass.

Thank you for the link - I’m going to give it a check now!

It does seem to be the case for me - I’ve seen far less overexcited comments, and far more questioning its present and future abilities to perform.

I think that at the bottom of the trough people will stop caring about it altogether except for functionality that has been proven to be useful for - such as a translation aid, coding aid, or to quickly pull off multiple images if you don’t mind overall quality.

I just generated a fresh example highlighting it. I’ll put it inside spoilers to avoid clutter:

In the “overexcitement peak”, I think that a lot of people were focusing on the correct output, and assuming that the rest would be solved “with enough elbow grease”, playing whack-a-mole with hallucinations. Now they’re noticing that it isn’t.