Lithnin @lithnin - Tumblr Blog

Posts

sociallyunacceptableart4

This was my art school’s water fountain. Drink from them wolf tiddies

mando-lore

Assignment misunderstood. I have now built a city.

mando-lore

Give it a day

alexaloraetheris

#laugh rule

sailorsally

Unpopular opinion but if you don't enjoy the process you should find a different thing to do.

And I think this is true in general but now I'm talking about it in the context of AI.

If you don't enjoy making art and only care about the end piece and how it'll look and how much traction it"lol get online then making art is not something for you, find something you enjoy from start to finish.

Same goes for writing: if you do not enjoy writing and rewriting and then some more and instead want AI to write for you, being a writer is not something you should pursue.

Sure, not every part of creative process is going to be equally enjoyable but you should get satisfaction from solving the problems along the way and you should get a sense of accomplishment on your way of "making the piece yours" and you should have a sense of ownership once you are done.

None of these things will come from typing in a prompt into chatGPT. And I am sad to see so many people are missing on the opportunity to experience the joy of making something with their own hands and brains.

theivorytowercrumbles

#this is so true#i know writers like to joke about hating writing#but like if you're serious?#if you actually hate the process of writing?#why is this what you've chosen to pursue#just do something else

kvothbloodless

Because I like creating stories! I like coming up with a plot and characters. I like worldbuilding. I like making decisions about themes and how I want to communicate them. I like figuring out pivotal scenes and how they'll play out. And while I don't love it, I also don't really have a problem with editing scenes and dialogue to fit better.

What I don't like, and indeed most days am not Able to do, is actually sitting down and converting all my well organized ideas and decisions and worldbuilding into tens of thousands of words of prose. I just do not enjoy it at all; I find it tedious and unengaging and it usually feels like pulling teeth. And even if I Did enjoy it, my disabilities make it near impossible to do anything like that most days.

And AI mostly solves that problem for me! Why Wouldn't I want to use it? Why shouldn't I use an awesome new tool that lets me find joy in creating stories I would not otherwise be able to create? Why shouldn't I want to share those stories with people?

I just don't understand tumblr user's immediate and outraged response to someone doing something that they also enjoy, just because they're doing it in a different way. I mean, I know a large part of it is because we like to equate suffering and struggle with value, but like. Cmon guys. Lets stop acting protestant (and ableist) here.

Something being harder to do doesn't make it inherently better! Hard work purely for the sake of hard work is dumb and helps nobody; you shouldn't be railing against people just because they're choosing to take advantage of a tool that allows them to produce art that they otherwise wouldn't. Or even art that they Would have made, but this lets them make it more easily! That's awesome! Being able to create a story or a piece of visual art in a day when it would otherwise take a week is amazing! What's the downside here? More art? More varied art (because now its easier to experiment and try new things)?

And yea, I will admit, AI isn't quite at the level where I can use it to create the stories I really want to Yet, its still so much better than any of the alternatives. So I ask again. Why Wouldn't I use it? Its letting me do something I enjoy, that I would not otherwise be able to do. Its not hurting anyone (please do not try to argue about how AI is killing the environment or stealing from artists unless you've actually done some reading and understand how the technology actually works). If its not for you, that's fine. Just like any tool, it won't be helpful for everyone. But that doesn't mean you should be hating on and ostracizing those who do find it helpful, nor that you should ignore any art produced using it.

swordandscytheandpen

Gonna be a little less polite and graceful than you, Ezra, sorry.

Notably absent from OP's post is any explanation or indication as to why one needs to "enjoy every step." It's all pure assertion, a cavalcade of moralizing "should" statements that only carry weight for those who are already convinced that AI developers are a witches' coven. Any particular reason a creator shouldn't focus on the end product? Any explanation as to why solving "the problems along the way" are so strictly required for accomplishment and authorship? No, OP's just dictating how other artists should and shouldn't engage with the creative process, which ironically has the effect of flattening the wonderfully diverse methods of artistic creation into a single orthodox Proper Way To Do Things. Artists, writers, musicians don't owe you a particular artistic process, they can and should use what creative modes and tools they see fit, up to and including LLMs, because choosing what resources to use, and how, is a creative decision too.

Guess what? I like writing, start to finish. For me, it's perhaps the only art form that is true for. I avoid using LLMs for creative-writing tasks because I enjoy having maximal influence over every step of the process. Yet I can recognize that this is not true, or indeed possible, for every writer. My own use case makes LLMs an unpreferable tool to use. That doesn't mean I begrudge the tool itself, or the authors who do choose to make use of that medium. I'm capable of recognizing that different artists have different methods, and that is a beautiful truth that feeds into the variety of works humankind is able to create.

Blanket opposition to LLM technology in creative contexts has no positive results. It's just another breed of purposeless reaction, and I have no idea why it's so widespread. You're policing people's means of expressing themselves, to nobody's benefit. OP, I'm glad that you get the most satisfaction from deeply engaging with the entire hands-on process; you've found the method that works for you. I do not always get that same satisfaction, and I can find first-drafting a bit tedious at times, but I do it anyway because that's how I can best forge the end product I desire. Some others wish to project their thoughts onto a canvas or page, but don't feel the need to stick their hands into every brushstroke or line of text to create a work that brings them joy and fulfillment. Why tell them how to make their art? You wouldn't do that with anyone else. Mind your own damn business.

satanfemme

the thing about LLM technology is it's a tool, it's non-sentient, it only does what it is you tell it to do same as any other tool. using generative ai is not a collaboration with another person any more than other artistic uses of tools are collaborations. LLM technology is not comparable to a tattoo artist any more than a paintbrush is comparable to a tattoo artist. comparing tool usage to collaborating with another person is a false equivalent that only makes sense if your are needlessly anthropomorphizing the tool.

tools have always been used in art, and which tools are used have always changed and adapted as society progresses. if someone is an artistic photographer, would you tell them "but YOU didn't make the actual piece, you just fed your ideas into a camera and are claiming the result as your creation"? (this was a genuine argument made at the advent of photography as an artistic medium btw. these reactionary claims are historically circular).

and on the other hand, if someone makes art with a paint set they bought from a store, rather than mixing the paint from scratch themselves, is that not also a Collaboration between themselves and the anonymized workers who crafted their supplies for low wages? why is this not generally considered a collaboration that voids the painter's claims to sole ownership over their piece; but using an LLM (which, again, is not sentient) is a collaboration an account of the LLM "doing the work for you" (which, again, is not really how it works).

the way people discuss ai-assisted art as "not really your own work" is out of step with how credit and ownership is otherwise assigned in the art world. there are interesting conversations that could be had here but "well it's not your work because you used a tool" isn't one of them.

lithnin

I continue to use the "apply the same reasoning to photography and see if it sounds like old-man-yelling-at-cloud insanity" test on anti-AI arguments, and it really helps separate out the few thoughtful critiques from a vast sea of reactionary kneejerking.

#ai art discourse

Thank you, OP of this thread, I have recently been experiencing an acute deficiency of completely insane takes and this topped me right up

#screenshot because OP took their ball and went home #uncharitable

alexanderwales

I spent some time today working on a little toy model game, wherein you have to arrange "ship parts" that are made of squares, paying attention to weight distribution, center of mass, aerodynamics, etc. It was a bunch of vibecoding to test whether the shipbuilding was actually a fun thing to do, something that wouldn't have been possible to do rapidly before the LLM, at least not with the current rust on my programming skills.

It's pretty fun, but would require a lot of work to add any amount of balance, especially when The Vision is that it would be a roguelite part-drafting FTL type of thing.

A few things that occurred to be during the prototyping:

It's fun to have two different modes for construction, "bolt" and "rebuild". When you're out and about, flying from planet to planet, you're using "bolt", attaching parts to your ship as well as you can, jettisoning things you no longer need to make space, trying not to unbalance it. But then you get to some kind of rest point, and you can take the whole ship apart to engineer it into the best version of itself. Like healing in Slay the Spire, but you're getting in the guts of the machine!

There's something really pleasing about polyominos, which is that they increase in complexity as you add in more squares. There are two shapes you can make with three squares, seven with four squares, and twelve with five squares.

It would be interesting to have the polyominoes have mechanical identity based on their shape. Storage is O-shaped, power is S-shaped, weapons systems are L-shaped, thrusters are T-shaped ... the problem with this approach is that if part of the gameplay is ship-balancing, we don't want there to be a "standard" ship. So probably three of the shapes are "core" predictable ones, cargo can be any shape, and then the other non-core functions? Mostly I like the idea of going from tertomino T-shape to pentomino T-shape.

Tiny little one-square or two-square shapes are good here as a method of balancing out the ship and making sure that aerodynamics are good? I'm thinking of them like little bonus pieces you're almost always happy to gobble up.

Shipbuilding will depend on tension, and if there's "aerodynamics", mass balancing, center-of-thrust vs. center-of-mass, then there can naturally be archetypes that arise from that, like aerodynamic "missions" or "events" or whatever, which you are either optimized for or not. All this would be for TBD gameplay, but I was working on the bones.

Center-of-mass, center-of-thrust, and aerodynamics are all somewhat unintuitive, particularly if the pieces have different weights to them, so this needs to be very clear in the UI when you're fiddling with ship design, super reactive to the player moving pieces around.

Aerodynamics is handled via rise-over-run along the left and right edges, and I don't know how to calculate that and show that calculation in a way that's instantly legible.

Once the bones are done, there are a lot of neat events you could do, things like adjacency buffs, debuffs that inflate a piece by a square, miniaturization that makes a piece smaller (but might unbalance the ship until reconfiguration), and stuff like that.

These are mostly notes for my own benefit, as I don't intend to throw away more time on this. Vibe coding and prototyping is nice to see the shape of a thing, but the actual work of game-making is a lot of testing, tuning, etc. To actually have a collection of rules for how ship parts are generated, and a long list of events, takes much more than minimum-viable design.

#somebody please make this

ninamiart

I wanted to quickly update Ankhe's outfit because I wasn't happy with her old ref sheet. I slightly simplified the dress and added a slit to make it more flowy.

#draeneiposting

Aeons from now, when humanity has fully mastered the laws of physics and bestrides the cosmos like unto the gods themselves, perhaps we will finally be able to develop the ultimate technology: a band-aid that doesn't fall off the moment it gets slightly wet

#this post brought to you by the blisters on my dominant hand #I've tried like 4 different brands and they ALL suck #venting

ozempicofficial

#laugh rule

dduane

The Star Trek title card generator is over here:

Make your own era-accurate Star Trek episode title cards in seconds. Boldly go!

The sitemaker (Josh Mayfield at https://bsky.app/profile/bean525.bsky.social) has asked for people to come in and kick its tires so he can work out any bugs they turn up. So do give him a hand, if you feel inclined. :)

(also, deeply amused by this one...)

thelaurenshippen

it's always "did you really need to burn down your workplace to prove a point" and never "how was the revenge arson? did you have fun doing revenge arson? were the flames pretty?"

thelaurenshippen

um.

partasah

🔥🔥🔥

Context for other users:

On the morning of April 7, 2026, a fire was started at the Kimberly-Clark Distribution Center in Ontario, California. It escalated to a six-alarm fire and took nearly twelve hours to extinguish. The facility was completely destroyed and declared a "total loss". There were no injuries or deaths. An employee who was working in the warehouse is accused of arson.

Authorities stated that the employee had shared videos of himself starting the fire on Facebook, saying "all you had to do was pay us enough to live" and "there goes your inventory" while lighting pallets of toilet paper ablaze.

ALL YOU HAD TO DO WAS PAY US ENOUGH TO LIVE

No injuries. No deaths. This was a protest fire. This is why unions are important, to prevent frustrations like these, to ensure people are paid enough to live.

🔥🔥🔥🔥

eddiepriestfuckingdiaz

No Injuries or Deaths

captainragtag

There were 20 people inside when the fire started, the fire being contained to a single building and not moving onto the very close residential neighborhood in famously flammable socal is thanks to the firefighters, not because this is a hero of the working class who strategically harms only corporations without death or injury to ordinary people, and the corporation will get their insurance payout and how many employees will be left entirely jobless? don't do this tumblr

lithnin

There is no such thing as "good arson", even if the person committing it has real grievances

twistedpictures

im off twelve steven universe percocet, i’m fumbling bisexual bitches left and right. i don’t give a shit. i had to smoke the grass i was too busy touching your waifu’s balls. I have precambrian sea kin memories i swim in Trilobite pussy. Puffed the Nagito pack so hard they’re flying pride flags half mast. this shit ain’t nothing to me man. Refer back to my bio motherfucker, check the pronouns you’ll see that I am him

#laugh rule #dracula flow

alexanderwales

I was playing through a hotel management game and accidentally trapped a guest in their room when it suddenly occurred to me that you surely you could mod in an H.H. Holmes murder hotel type of thing.

Then I went to go search Steam to see whether anyone had in fact done this, or made an H.H. Holmes game where you build a death trap, and apparently no one has been that tasteless.

(I did release the accidentally trapped guest.)

alexanderwales

Alright, here's my pitch for a Murder Hotel game:

You are a serial killer running a murder hotel (or in the fashion of hotel games, running a series of murder hotels). You have two main goals: murder people and evade detection. The win state for any given hotel is amassing money, and the lose state is bankruptcy and/or being caught by the police.

Every person who works in your hotel, is contracted, or stays there, has a suspicion meter that slowly rises over time when they notice something is wrong. This can either be noticing that the layout is weird, seeing a suspicious stain, or The Smell, which I envision as a major mechanic. If suspicion gets high enough, they'll either start making investigations of the people or the property, or much worse, bring in the (disorganized, incompetent) police. So suspicion management is one of the major things you're doing, either assuaging concerns, raising pay enough for them to forget about it, or firing employees you think might be onto you. Alternately, if someone has high suspicion, you can just kill them.

Separately, everyone has connection, a measure of how much they'll be missed by their friends and families. Ideally, you want to be killing only people with low connection, which is unlikely to attract unwanted attention. Kill someone with too high of connection and you might get a private investigator on you, or less threatening, a loved one snooping around your murder hotel.

You are the Proprietor, a sociopathic but intelligent and charming hotelier. You have a physical avatar that needs to balance 1) lowering suspicion through charm 2) murdering people 3) disposing of bodies. If you have no employees, you fill any role in the hotel, including some basic construction work. Unfortunately, you have The Hunger, a desire to kill, which must be periodically sated. If you don't sate The Hunger, then you'll lose control of the Proprietor, who is liable to kill in his own messy way, which you'll then have to clean up, and is likely suboptimal, especially if he decides to kill a valued employee, or do the kill in a more public place, or while police are visiting. The Hunger imposes a time limit, as it grows over time, requiring more kills, splashier kills, etc.

So the game is one of tensions, trying to manage all the people, delegating tasks that are going to necessarily raise suspicion, cycling through guests and staff, quietly getting away with killing for as long as possible, managing cash flow. To this end, you hire on rotating construction crews to build different parts of the murder hotel: soundproof rooms, airtight rooms with gas lines, secret passages, chutes leading to the basement, vats of acid to dissolve bodies, etc.

And in the classic style of tycoon games, you beat a map and then move on to another, but here you have a reason to do that, since obviously sticking around in one place for too long would greatly increase your chances of getting caught.

#I would play this #game design