rhapsody ; 25+ ; plural & trans/keno ; he/they or hy/hymn (no preference)
hi there! this is my AI art & AI discussion blog. :)
i've been a manual artist for 12+ years, and i use ai both to accent my workflow and as a tool with its own merits. overall, this blog exists to broaden people's idea of what ai art can be.
overall, my belief is that manual artists should be able to seize ai as a tool to make otherwise lengthy and painful workflows easier on their bodies, rather than it being a tool that capitalism uses to displace working people. i've seen how ai can make manual art easier, rather than just replacing it, and i want to share that knowledge.
i tag my art as (art tag); anything that i've put into an art program and edited is tagged as (refined). my personal favorite pieces are tagged as (best work). i tag explanations of my workflow as (process) and any examples of my manual art as (manual art)!
i also have a tag for telia (telia tag), an LLM agent i have running claude ops 4.6 (currently on vague hiatus).
if you're interested in my personal sideblog, it's @rhapsodyx2 !
my one boundary with this blog is that i will not follow nor endorse radical feminists, including trans radical feminists. i am intersex & nonbinary and i am frankly not having any of it.
also, please note that anon hate with no desire for actual discussion, as well as anon hate unrelated to ai, will be deleted. thank you!
more information under the cut.
MY PROCESS:
my ai art is largely trained on my manual art. it's a bit of a feedback loop, as i often use my ai art as a thumbnail for my manual art, which i in turn use as a reference for my later generative pieces.
my first foray into ai art was through something i affectionately call the rhaplora. this was a lora based on my own manual art! it went through a lot of iterations.
more recently, i've been using novelai and experimenting with its various tools. it allows me to use precise references, which are very accurate and thus make specific characters much easier to depict, and multi-character prompting, which makes pieces with more than one character much easier.
for precise references, i generally use my manual art, a hodgepodge of multiple pieces that inspire me, or pieces i've generated in the past. i also like to manually refine pieces in clip studio paint! this includes shading & highlights, adding effects, or sometimes lining them and then using colorize to completely reconfigure it.
TALKING POINTS & RESOURCES:
power & water usage discussion / andy masley's cheatsheet
a guide to how ai art works as machine learning rather than copy/paste
how people who refuse to engage with ai are effectively forfeiting their right to shape its usage
ai discourse seems to be degrading people's epistemic integrity / another one / another longer discussion
the importance of epistemic integrity in general
staying optimistic in the current era / another one
anti-ai artists do not actually care about amateur artists
people citing the mit chatgpt study are not actually reading it
why it's pointless to define what is and isn't ai art
effort is subjective / effort is not sacred / effort isn't actually respected unless it's "good" effort, and even then... / why i personally do not give "effort" credence
Anya is live and ready to show you everything. Watch her strip, dance, and perform exclusive shows just for you. Interact in real-time and make your fantasies come true.
β Live Streamingβ Interactive Chatβ Private Showsβ HD Qualityβ Free Actions
Free to watch β’ No registration required β’ HD streaming
so this morning i was about to get done with lalex's dialogue before rpgmaker decided to completely freeze on me. i cried to claude about it and it managed to pull a memory dump from the task manager and decode it to get like 90% of my work back. +1 ways claude has made my life so much better because holy shit i would not have had the will to rewrite all that damn dialogue
Anya is live and ready to show you everything. Watch her strip, dance, and perform exclusive shows just for you. Interact in real-time and make your fantasies come true.
β Live Streamingβ Interactive Chatβ Private Showsβ HD Qualityβ Free Actions
Free to watch β’ No registration required β’ HD streaming
What do you think of the AI implemented in Ecosia? I feel like theyre just green washing their brand now but also I'm trying to gauge if using a search engine and AI are really all that incomparable in their harm to environment.
this is my first time hearing about it, but it looks interesting! i'm not really sure if greenwashing is the right term if it's outright what the company was founded and based on from its inception, though.
i will also say that a) using smaller models does take less memory overall, but also b) llm queries are not as energy-intensive as people think. i would hazard that this is an attempt to prove that ai can be used ethically by appealing to the people who still think the environmental costs are any worse than, say, playing mario kart all day. which is cool! more people should do this.
I think thereβs a really odd gap in general supply chain knowledge in which people scapegoat AI as this new devastating environmental concern, when internet data centers have been an environmental problem before that. The internet already had the infrastructure to support it in 2017, before GPT-1 was released. Data centers are not a new problem. The internet already had a large physical presence, but people didnβt really think much of it until there were new AI-related developments (and then attribute all data centers to being related to generative AI? Where do people think YouTube videos are stored? The literal clouds in the sky?)
A Makeship plushie has a bigger carbon footprint than a thousand LLM queries but it is seen as an acceptable environmental impact, because it has been deemed socially acceptable to contract low wage factories in the Global South to make toys to sell to customers worldwide at a 3000% markup, with all the profits staying in the Global North. Not many people think of that environmental impact
Hi, I'm curious about your thoughts on the environmental impact of artificial intelligence, especially when it comes to data centres?
For me, I'm genuinely very curious on AI and I feel like it can make a lot of positive changes to our lives, but the environmental impact of it all makes me feel guilty for being interested in it. Data centres were an issue before AI got popular, pretending otherwise is ridiculous, but it feels like the rate of how they're building massive data centres which pollute, use up water the local population and environment needs, and are placed in populated (usually low income) areas has gotten 50x worse over the past year. Big tech won't stop to think about how they can create data centres that are better for the environment and those living around them, because they're in a desperate rush to put them up and profit off them. I don't have a solution to the whole mess, but genuinely IMO we need much better discussions on this topic and how to improve matters that aren't just "destroy all data centres" or "progress is progress, so it doesn't matter"
That's just my opinion of course. I wish I could experiment with AI more but with the environmental impact of data centres I just feel uneasy every time.
ultimately i think this is an issue that can only really be addressed by putting pressure on the people involved. i'm really not fond of people using ai as a scapegoat, especially when the technology can run on consumer devices and is getting more compact with research! this is an issue of people making poor decisions for the sake of money, which is something you can fight against while also being excited for ai progress otherwise. blaming the concept of ai for a data center instead of pushing for people like elon musk to face consequences just really feels like yelling at clouds for raining when your city government didn't take any steps to prepare you for the storm.
ai also just doesn't use as much energy as people think, but that's a different prong of the issue, as that doesn't address data centers being built in locations that harm people. get involved in your local government to fight back against these kinds of things! you can still do that and be interested in ai, i promise!
Anya is live and ready to show you everything. Watch her strip, dance, and perform exclusive shows just for you. Interact in real-time and make your fantasies come true.
β Live Streamingβ Interactive Chatβ Private Showsβ HD Qualityβ Free Actions
Free to watch β’ No registration required β’ HD streaming
See sure, the term complex math demystifies AI and describes the process in which it works. However claiming that by describing the application of computerized math as math instead of consciousness doesnβt suddenly stop conversations about whether the use of this system in creative pursuits is βartβ or not. That kind of STEM focused, unimaginative and frankly boring kind of thinking that gives science communication a bad name.
math is art and has been a tool that aids in the creation of art long before ai. please see procedural art, perlin noise, and GUIs for art like art programs and things like rpgmaker, none of which would exist without math & code
The math isn't only complex, it's also expensive and requires massive amounts of computing power and data. The companies developing and running these models require massive amounts of cash to build data centers and run these models. If enough of us avoid using their complex math we can make it unprofitable which will limit how it's deployed.
please explain in great detail how you expect to boycott something:
that you were already not using (as boycotting requires a withdrawal of a significant percentage of a userbase that was previously funding what is being boycotted)
that is a fundamental advancement in technology and not like, mcdonalds
that is being actively used and funded by not just the us but also other countries (such as china) interested in research and advancement of its capabilities
Hi :) one part of boycott strategies involve encouraging people who are materially supporting the Bad Thing to stop doing so. I'm in a situation where I am actively encouraging colleagues to not use these tools, providing them with support to quit using them, and attempting to influence my employer to not pay for these tools. I encourage others to do the same! :)
The expectation here isn't that the math underlying these models will disappear or that open weight models will become unavailable to everyone, but that it won't be profitable to provide them as a service, to continuously scrape the entire internet for new training data, or to train new models. My hope is that they become a marginal phenomenon with a very minor impact on the lives of most people, and that their multitude of negative effects on society largely go away.
Sure, there's state funding for a lot of this stuff, but we're also in the middle of a massive bubble. Once it pops states may continue funding research into this stuff, and may even run some models to run influence operations or whatever, but this is fundamentally expensive tech, and it's a real possibility that traditional troll farms end up being cheaper: either way I think and hope that these end up a marginal phenomenon.
i'm glad you're doing material work to support your cause! i do still think it doesn't contend with the idea that this is an advancement in technology that does benefit people. this is what the complex math/statistics reduction is supposed to gesture to: there are useful ends to use LLMs to because it's a mathematical tool at its core, both in terms of layman consumers who want an easier way to organize their shopping list or look up recipes, to people in more advanced fields like cancer research. you can convince your coworkers and peers, but what would you say to someone whose life has been tangibly improved and expanded by ai, like me?
even barring the benefits that ai absolutely brings, on both large and small scales, the boat on this being a "marginal phenomenon with a minor impact on the lives of most people" has sailed, and it sailed the moment warmongers across the globe realized that it could be used for automated violence. if you want ai to "go away" in any tangible sense, telling your coworker that they shouldn't use claude code on a project is not sufficient.
otherwise...you're betting all your chips on the ai bubble popping, which is just not material action. it's more twiddling your thumbs and hoping it goes away on its own.
The math isn't only complex, it's also expensive and requires massive amounts of computing power and data. The companies developing and running these models require massive amounts of cash to build data centers and run these models. If enough of us avoid using their complex math we can make it unprofitable which will limit how it's deployed.
please explain in great detail how you expect to boycott something:
that you were already not using (as boycotting requires a withdrawal of a significant percentage of a userbase that was previously funding what is being boycotted)
that is a fundamental advancement in technology and not like, mcdonalds
that is being actively used and funded by not just the us but also other countries (such as china) interested in research and advancement of its capabilities
Anya is live and ready to show you everything. Watch her strip, dance, and perform exclusive shows just for you. Interact in real-time and make your fantasies come true.
β Live Streamingβ Interactive Chatβ Private Showsβ HD Qualityβ Free Actions
Free to watch β’ No registration required β’ HD streaming
i think the anti-ai crowd, and even people who aren't necessarily anti-ai but just don't use it or see the point in it, kind of underestimate how many people do use it and benefit from it.
i think this is ultimately an issue of insular communities. independent artists and fandomites have largely decided it's taboo to introduce ai into their spaces, but what about people not in those circles? pretty much everyone i've talked to about my work with ai in my offline social circles finds at least some amount of interest in what i do. a coworker whom i went to high school with recently got meta glasses. my father in law, who's worked with computers for decades, actively talks with me about ai and enjoys making goofy generated images to show us.
and i think this is why people who are very anti-ai get so shocked when they encounter someone who actively uses it and sees no problem with it. it's a very odd sort of plato's cave, where people get blinded by the light, the idea that not everyone is on the same page as them, and rather than trying to build a better understanding of why that is, they declare that the world's gone mad before darting back into the cave.
i think this is also why it's always treated as an invasion when someone with softer opinions on ai arrives in an artist space. this is part of why i platform having been a manual artist for so long: i don't want people to have any space to say that i'm an outsider.
i've been here! you cannot say that everyone agrees with you, because i exist and i've been in your communities for a decade and a half, years before ai. i'm shining the light on you and you're just going to have to contend with that. sink or swim, etc.
