Squid🦑 @squi.dev · 03/10/2026a games code is also just as much a creative work as the rest of the project (and imo code in general even outside of games) 040
Reposted by Squid🦑LJ Amber with false fangs and little bat wings @ljamberfantasy.com · 03/10/2026Sorry, using 'AI' 'Agents' for coding on your game counts just as much as generating visual or audio assets using an LLM. You're still stealing the labour of human beings (writers and programmers) to train a model to make human beings redundant. 3132
Squid🦑 @squi.dev · 24/09/2026waow the people have spoken and there is at least a few people, maybe even four or five, want this new and amazing chat platform beyond most mortals comprehension. tempted to make it. 010
Squid🦑 @squi.dev · 24/09/2026who here would use an SSH based bulletin board system in this year of our suffering 2026 if i made one? please sound of, the two or three of you 210
Squid🦑 @squi.dev · 18/09/2026i am highly unhappy with how there is the 'Harry Potter fanfic sex cult' and how often it id relevant to major political topics! 022
Squid🦑 @squi.dev · 10/09/2026[impossible difficulty] do I choose some guy who has a thing subscription he may be willing to share or all my friends mhhhh how could i possibly make the right choice here, so difficult! 010
Squid🦑 @squi.dev · 10/09/2026could be worse, could be your draft avout that play with a yellow king ... 110
Squid🦑 @squi.dev · 10/09/2026id actually prefer they dont do that, its like as if i liked people getting into chemistry through a drug addiction 010
Reposted by Squid🦑Fallible Things @fallible.net · 09/09/2026Nausea blog.fallible.net/it-breaks-a-...blog.fallible.netIt Breaks a Village: Bevy's 6th BirthdayThis post, like all my posts, was written by me. It came out of a lot of conversation and reflection with many people. It is long. I promise you there's valu... 03420
Squid🦑 @squi.dev · 09/09/2026i kinda wish i could be more optimisitc about AI. but like, i look at the people behind all this, and i end up with nothing but pessimism for the tech. 000
Squid🦑 @squi.dev · 09/09/2026y'know i used to think 'woah AI cool' and than these last few years happened and now i want to... idk but i want to do something 110
Squid🦑 @squi.dev · 08/09/2026me after recruiting @sachy.dev for my dark and evil endeavorsstatic.klipy.comSilent Witch Monica EverettAlt: Monica Everet, from the anime silent witch, chuckling adorably and happily 010
Squid🦑 @squi.dev · 08/09/2026god i love shitposting on the internet. i wish they paid me for this shit but alas, the suffering of modern life /j 021
Squid🦑 @squi.dev · 08/09/2026you know its kinda painful to watch what was once a fun and great community very rapidly imploding, deleting the memes channel, and doing jack shit to solve the problems that are actual there. i mean i guess you dont need a meme channel if you're a clown 000
Squid🦑 @squi.dev · 01/09/2026Mhh, I love economic rent seeking... Which is like the one think thats universally considered economically harmful, but hey, would be a bit much for a goverment to regulate them... 010
Squid🦑 @squi.dev · 31/08/2026.....i have a system with a RTX 5070... and it just works... on wayland... on archlinux... 110
Squid🦑 @squi.dev · 27/08/2026sophie, it resolces for everyone else, this is an issue with _your_ dns not theirs 110
Squid🦑 @squi.dev · 25/08/2026some real unit gore as well, token/s/kW, also know as.... token/kJ... but i guess the units are normal rative to the graph gore, like, waow, 100x performance in... tokens/J at a set model at set output rate? how does that measure even relate to anything? 010
Squid🦑 @squi.dev · 19/08/2026Like, fuck those companies and their empirically provable bad tools, that are doing actual harm, but also don't make bullshit claims about things. You can point at the example and go "they clearly cant do it," and you'd be right. 000
Squid🦑 @squi.dev · 19/08/2026The entire claim that doing this were somehow incompatible with those laws just... doesn't follow out of an already incorrect argument about how the models even work, or how detection of them works. Its just one layer of wrong on another layer of wrong on another layer of wrong. 100
Squid🦑 @squi.dev · 19/08/2026(doubling capabilities takes like, several orders of magnitude more compute, lmao, because this does indeed not break thermodynamics, and thus costs a lot of energy!) 110
Squid🦑 @squi.dev · 19/08/2026And hey, we happen to provably know that neural networks are in fact capable universal function approximation, if just made big enough. If you make it big enough, and by extension stupidly expensive to run, it can do the task. The cost is just _slightly_ non linear, though. 110
Squid🦑 @squi.dev · 19/08/2026Especially if your argument is so unsound that it basically implies you can't train a neural network at all. Like, that whole 'going to infinity' thing makes no sense. Nothing is going to infinite, the models error rate (in the example of meaning output text not looking human) goes to zero. 100
Squid🦑 @squi.dev · 19/08/2026Sure, most of those companies are just in for a quick buck and don't give a fuck if their tools are accurate, but to claim that some law of nature were violated by such a machine existing is just nonsense. 100
Squid🦑 @squi.dev · 19/08/2026So... they fine tune the model to leave traces. That's what Anthropic just started doing. And with it done _intentionally_ it _absolutely_ going to be possible to detect. And it probably is on other models too. 100
Squid🦑 @squi.dev · 19/08/2026In general, things the LLM doesn't need to do very well, or doesn't need to focus on all that much. And well, the people training these models do actually have a self interest in keeping them distinguishable. Feeding older AI slop is, in fact, not great for making a newer slopinator. 100
Squid🦑 @squi.dev · 19/08/2026Again, they are NOT made to be human indistinguishable, but achieve some human made reward metric. This is, in big part, because a whole lot of human text is not a good base! A LOT of the raw text is nonsensical arguments, for example. Or just people saying hello, or posting about their vacation. 100
Squid🦑 @squi.dev · 19/08/2026The actual technical question with LLM detection feasibility now is if those profiles share certain properties that human profiles do not. The answer is, probably yes for old models, and absolutely for most new ones, because AI labs intentionally fine tune them for it. 100
Squid🦑 @squi.dev · 19/08/2026The thing that makes LLM detection hard now, is not some inherent impossibility, but the models tendencies to change this linguistic 'profile' based on user input. So if you throw a prompt at it, like 'write in x style' it will often adopt another, distinct 'voice.' 100
Squid🦑 @squi.dev · 19/08/2026One example is that you can analyze the usage of location specific slang or terminology. People have accents, and not just in their speech! As a result, with enough text, it is very much possible to specifically identify the author of something! 100
Squid🦑 @squi.dev · 19/08/2026Humans do not write text that is the exact average. Humans write texts with statistically significant differences. In fact, you can compare texts to find if they have the same author! This is not a new thing either! 100
Squid🦑 @squi.dev · 19/08/2026Also, the thing is that, a text or any other thing, that has all the average values of a large diverse population, is very strange! A text that is statistically closest to the overall dataset average of all human text, is not very human look at all! 100
Squid🦑 @squi.dev · 19/08/2026And compute, famously, cost energy! Which is why this doesn't break thermodynamics. It also doesn't break information theory, because we already know what human text looks like, and are just extracting out the patterns that make it look like so via computation. 120
Squid🦑 @squi.dev · 19/08/2026And if we assume that was possible, and that such a detector could be used to train better models, which in turn would lead to better detectors, that would not break any laws of information theory of thermodynamics either! That would still require immersive amounts of compute! 100
Squid🦑 @squi.dev · 19/08/2026How would a machine capable of detecting that difference be usable to train a 'better' model? Only _if_ the goal was to make it indistinguishable from human, which it is not, the goal is to make it succeed at a set of training tasks! Tasks that in fact don't lead to it being all that human like. 100
Squid🦑 @squi.dev · 19/08/2026The model does not produce statistically indistinguishable from human outputs, but the outputs that satisfy its training rewards system most strongly. Those are very different things. And the rest of the argument is less logically. 100
Squid🦑 @squi.dev · 19/08/2026That is not how any of this works. This is not even close to why its hard to make a machine like that (and very much possible to do so, in theory). 100
Squid🦑 @squi.dev · 17/08/2026it got turned into a 260mm doll by a malevolent friend. it regrets its doll decisions leadinf up to this moment. 000
Squid🦑 @squi.dev · 15/08/2026bsky is better the twitter ever was, because i can have the full power load of posts blastes onto my phone at m leisure jestream goea brrrr 010
Squid🦑 @squi.dev · 13/08/2026when you suddenly realise you didnt eat properly in like three days and that maybeee thats why you are doing badly..... anyway food time!! 110
Squid🦑 @squi.dev · 13/08/2026As has recently made some rounds on social media, in german we would refer to this as 'hobbyloses Verhalten' or hobby less behavior, which is to say this person really needs to get something to do. Like maybe pick up a pen and do some art... Wouldn't that be a crazy idea... 041
Squid🦑 @squi.dev · 13/08/2026Also, that implied idea that they somehow think this is beneficial to their cause of normalizing 'AI art.' Being an awful person is very rarely good for convincing people. (Something that everyone needs a reminder on though from time to time, for what its worth.) 141
Squid🦑 @squi.dev · 13/08/2026Is this some sort of absurd 'I'm better then you' power play or just spite? Cool, you have a dataset, of data you know you have explicit non-permission to do something with, and you're going to do what? This is a really deep case of 'owning the libs' syndrome. Just... shooting yourself in the foot? 141
Squid🦑 @squi.dev · 13/08/2026This is some of the weirdest, most just hate fueled things I've seen in a while. Like, 'DefendingAiArt' by actively threatening to do... what? Train a model on it just to be a awful person? Just, what does this person gain from this? 1133
