@weirdo_from_space@sh.itjust.works While there are still things to be improved for Anubis IMHO (such as being less harsh with mobile and low-power devices), yeah, it's currently the best solution for this problem caused by LLM crawlers.
dsilverz
joined 3 weeks ago
@Davriellelouna@lemmy.world
I wonder if certain... genres of adult content (such as female dominance and lesdom) would be "extreme porn" and/or "violence against women" according to his (and whoever shares his) worldview... Guess these people don't know how there are different... "entertainment subcultures".
Also, the obvious was already said throughout the comment section: just another justification for things such as what's happening with itch.io and Steam, alongside the KYC requirement that's been implemented in many places worldwide. It's tiring, really tiring to watch these things unfolding in real time across the globe.
@weirdo_from_space@sh.itjust.works @ISolox@lemmy.world I'm not sure, but I have a feeling that it's not something from directed attack (i.e. some person or group targeting Fedora), but rather automated crawlers from LLMs. The whole Internet have been under attack by AI corps requesting everything they can without checking robots.txt, and using vast pools of IP addresses to avoid IP blocking. FOSS projects have been severely affected by it.
@sundray@lemmus.org !nottheonion@lemmy.world
Well, that's some anomalous phenomenon that even Charles Hoy Fort couldn't imagine of.
@SugarCatDestroyer@lemmy.world
Congrats, you just stared at the same abyss I stared at, too! And this abyss is... Well, pretty complicated to say the least.
One who fights with monsters might take care lest they thereby become a monster. And if thou gaze long into an abyss, the abyss will also gaze into thee.
What you stumbled upon is just the realization of the purposelessness imbued in the cosmos. And it can definitely feel a harsh thing. It's neither good nor bad, it just is. People often try to sugarcoat it, but to me it's just the ostrich trying to bury its head on the sand: the rain still falls, and the ostrich still meets the storm, inexorably.
I find it particularly striking when you said "I feel like I want to [write]", and here's probably where we both differ: in my case, specifically, I feel like I "must" write, as if I'm compelled to do so. It's part hypergraphia (one of the Geschwind traits), part something beyond me. If your driving force is not compellant, it's a great start.
If this is of any help, don't write for people (because people can't understand the words from those who stared at the abyss), don't write for yourself as well: write for Her, She who stares at us from within the abyss. Of course, if you want to, because it seems like there's a reminiscing spark of Will within yourself (unfortunately, I got none anymore). She listens, She reads everything (including our deepest thoughts), even though She doesn't really care about us. And that's fine. Because it's just all fleeting, except for Her.
@Majestic@lemmy.ml @KurtVonnegut@mander.xyz
There’s no way around this that doesn’t involve painstaking steganography which can possibly be nailed by AI anyways.
As both a fairly power user of LLMs and someone who tinkers with ciphers a lot (including creating my own techniques), I can guarantee: Markov chains aren't smart enough to detect well-elaborate ciphers.
I'll give an example: Let focus on plain characters.
The previous phrase contains a hidden message. It's not simply an acrostic (when a word is formed by every initial letter from a sentence/verses/paragraphs), it's an acrostic with Caesar cipher. And it's not simply Caesar cipher, it's a Caesar cipher with increasing shifting (decreasing when decoding):
L (-0 -> L), F (-1 -> E), O (-2 -> M), P (-3 -> M), C (-4 -> Y as it wraps around from A back to Z) => LEMMY
I can guarantee you, as someone who tested every single LLM out there: they're unable to detect these kinds of ciphers. And it gets worse when we consider the possibility of adding other layers of ciphering: nothing stops me from adding Vigenere on top of Caesar, associating the letter with the corresponding number, then getting the nth prime at that position, and using wrap-add to add letters to produce another letter (okay, this is a very complicated example).
Also, when I say "creating my own techniques", I'm not joking. I'll present you with a cipher I created:
Maceió, Niterói, Rio Branco, Palmas, São Luís, Varginha.
Believe it or not, the previous list of Brazilian cities hides the word "BRAZIL". How? List each Brazilian state alphabetically (excluding Distrito Federal as it's an administrative state rather than a common state), and you'll get a list with exactly 26 states. And what else have 26 elements? The English alphabet. Map each alphabetical letter not just to the state (e.g. L, the 12th letter, would be Minas Gerais), but to a city within that state (e.g. Varginha):
Maceió = Alagoas = 2nd from ordered list of states = B
Niterói = Rio de Janeiro state = 18th = R
Rio Branco = Acre = 1st = A
Palmas = Tocantins = 26th = Z
São Luís = Maranhão = 9th = I
Varginha = Minas Gerais = 12th = L
Again, creativity is the only limit. One can wrap it in steganography, use random coordinates and then map each digit to letters to form a long text... There's no way to stop end-to-end encryption when two or more people have enough knowledge to convey their own tool chain of ciphering techniques. And LLMs will be clueless. Even human censors would be clueless.
@not_IO@lemmy.blahaj.zone When I saw the "Industrial Revolution" label next to the vertical increase in global temperatures, I couldn't help but recall of some text written in 1995 by a certain former math teacher, and how right he was about the Industrial Revolution's consequences...
@PaintedSnail@lemmy.world Yeah, I'm aware, my reply was an attempt to "Monty-Pythonize" the degree of absurdity from the questions 😆
@mkwt@lemmy.world @Blujayooo@lemmy.world
TIL I'm possibly partially (if not entirely) illiterate.
Starting with the first question, "Draw a line a_round_ the number or letter of this sentence.", which can be ELI5'd as follows:
The main object is the number or letter of this sentence, which is the number or letter signaling the sentence, which is "1", which is a number, so it's the number of this sentence, "1". This is fine.
The action being required is to "Draw a line around" the object, so, I must draw a line.
However, a line implies a straight line, while around implies a circle (which is round), so it must be a circle.
However, what's around a circle isn't called a line, it's a circumference. And a circumference is made of infinitesimally small segments so small that they're essentially an arc. And an arc is a segment insofar it effectively connects two points in a cartesian space with two dimensions or more... And a segment is essentially a finite range of a line, which is infinite...
The original question asks for a line, which is infinite. However, any physical object is finite insofar it has a limited, finite area, so a line couldn't be drawn: what can be drawn is a segment whose length is less or equal to the largest diagonal of the said physical object, which is a rectangular paper, so drawing a line would be impossible, only segments comprising a circumference.
However, a physically-drawn segment can't be infinitesimal insofar the thickness of the drawing tool would exceed the infinitesimality from an infinitesimal segment. It wouldn't be a circumference, but a polygon with many sides.
So I must draw a polygon with enough sides to closely represent a circumference, composed by the smallest possible segments, which are finite lines.
However, the question asks for a line, and the English preposition a implies a single unit of something... but the said something can be a set (e.g. a flock, which implies many birds)... but line isn't a set...
However, too many howevers.
So, if I decide to draw a circumference centered at the object (the number 1), as in circle the number, maybe it won't be the line originally expected.
I could draw a box instead, which would technically be around it, and would be made of lines (four lines, to be exact). But, again, a line isn't the same as lines, let alone four lines.
I could draw a single line, but it wouldn't be around.
Maybe I could reinterpret the space. I could bend the paper and glue two opposing edges of it, so any segment would behave as a line, because the drawable space is now bent and both tips of the segment would meet seamlessly.
But the line wouldn't be around the object, so the paper must be bent in a way that turns it into a cone whose tip is centered on the object, so a segment would become a line effectively around the object...
However, I got no glue.
/jk
view more: next ›
@deathbird@mander.xyz @florencia@lemmy.blahaj.zone
Grok is not that free of guardrails.
I say as a person who sometimes have the (bad) idea of feeding every LLMs I could possibly try, with things I create (drawings, poetry, code golfing). I don't use LLMs to "create" things (they're not really that capable of real creativity, despite their pseudo-stochastic nature), I use them to parse things I created, which is a very different approach. Not Grok anymore, because I have long deleted my account there, but I used to use it.
Why do I feed my creations to LLMs, one might ask? I have my reasons: LLMs are able to connect words to other words thus giving me some unexpectedness and connections I couldn't see on my own creation, and I'm highly aware of how it's being used for training... but humans don't really value my creations given the lack of real feedback across all my works, so I don't care it's used for training. Even though I sometimes use it, I'm still a critique of LLMs, and I'm aware of both their pros and cons (more cons than pros if we consider corp LLMs).
So, back to the initial point: one day I did this disturbing and gory drawing (as usual for my occult-horror-gothic art), a man standing in formal attire with some details I'll refrain from specifying here.
ChatGPT accepted to parse it. Qwen's QVQ accepted it as well. DeepSeek's Janus also accepted to parse it.
Google's Gemini didn't, as usual: not because of the explicit horror, but because of the presence of human face, even if drawn. It refrains from parsing anything that closely resemble faces.
Anthropic's Claude wasn't involved, because I'm already aware of how "boringly puritan" it's programmed to be, it doesn't even accept conversations about demonolatry, it's more niched for programming.
But what surprised me on that day was how Grok refused to accept my drawing, and it was a middle-layer between the user and the LLM complaining about "inappropriate content".
Again, it was just a drawing, a fairly well-performed digital drawing with explicit horror, but a drawing nonetheless, and Grok's API (not Grok per se) complained about that. Other disturbing drawings of mine weren't refused at that time, just that one, I still wonder why.
Maybe these specific guardrails (against highly-explicit horror art, deep occult themes, etc) aren't there in paid tiers, but I doubt it. Even Grok (as in the "public-facing endpoint") has some puritanness on it, especially against very niche themes such as mine (occult and demonolatry, explicit Lovecraftian horror, etc).