@stefan Telling the world about the things I care about is my motivation as well. I want to learn how to write well so I gain another venue for expressing my thoughts in a way that's different (and would be impossible) through my art. I don't plant to write full-fledged stories, though.
I really hate my elementary school teacher for ruining writing for me. I read a lot of books and the essays I wrote back then were so good that I couldn't believe I wrote them when i found them a few years ago. But because my teacher didn't like me and gave me bad grades regardless, I stopped putting in effort. Fucking asshole.
@coffee I hate what social media has done to public discourse so that everything is turned into a black-and-white issue. Reality is never that simple. I can use an LLM and still think that corpos must be held accountable for shitting all over the concept of consent.
@sergaderg Now I'm curious what really needs more power: training one model, or millions of people running it. It's obvious that one person running the model needs way less power than training, but running happens multiple times, and is done by multiple people, whereas training happens once.
I like to think that I'm handling LLMs as ethically as I can. I run it on my own computer, which consists of hardware chosen to not require a shitload of power. I even got a workstation GPU because they have better performance per watt and it was the only upgrade path that wasn't more of a "kinda upgrade-ish". (Newer GPUs all need more power, so I had to look at weaker cards of newer gens of consumer GPUs). I had to do that because this house is old and the wiring sucks, so the power trips pretty easily.
I probably sound overly defensive about my LLM experiments because I know how poisoned AI as a topic is. I worry people will see my posts and go "ew, he's going to add slop to his website" and unfollow or block me. I know not everyone will, but it makes me think twice whether I want to post about it.
@lukadjo I don't rewrite anything because I simply don't take any of its outputs for my post. I just generate and observe and draw lessons from it (the parts that don't suck, because they certainly exist). I'm still going to write my post myself.
One of them was having it generate full essays in one go after supplying a shitload of background info, and modify the request to get different writing styles or structures. I thought it would give me ideas on how to approach my post, but the pacing tends to be very off and focuses on the wrong things.
My current method is providing less background info and having it write just an opener. The quality with this method is way better, and it's easier to observe how the writing changes depending on how I change my request. This helps me see how I could improve my structure and phrasing to be more varied and interesting, and it also gives me new ideas I hadn't considered before.
Using an LLM for learning to improve my writing works well with my learning process. I never followed tutorials or advice directly because it doesn't feel like I gain a true understanding of whatever it is I want to learn. Instead, I just play around with the possibilities I have, copy what other people have done and adapt them to my own ideas. I gradually learn how to do things on my own. That's how I learned programming and drawing vector art.
I think it's better to talk about my experiments with local LLMs here. I'll be playing around with Gemma 3 quite a bit since it's the first time a local LLM can probably help me with my writing. As an assistant, not as the writer. I still want my texts to be my works.
The first time I saw how my understanding of identity is different was when I first talked with someone who was a "therian" as well (but didn't use the label for themself). At that time, I was still feeling closer to foxes and mentioned how being one was part of my identity. They responded that it was their identity, not just a part of it. That seems to be true for a lot of people, as far as I can tell.
I could never understand it because there's so much more that makes me "me", independent from the desire to be an animal. Like I said above, I love drawing and programming, I listen to certain types of music and watch anime, I'm quiet and introspective, I'm asexual and don't care about gender, I hate standing out, and so on. There's just one identity, and that's "me" — the sum of all these things.
Some more thoughts about 2 topics I saw as reactions to my recent posts:
--
The first one is that my 2 modes of being online may resemble plurality.
I don't consider plurality a good approach to handle my own identity, though. People always are in different "modes" where their behavior and mindset differ. You'll seem like a different person around your coworkers compared to when being around your friends. As far as I'm concerned, my modes are just that and I see no reason to compartmentalize each one as their own entity.
--
The second topic is about one's identification with humanity.
I said last time that I consider myself to be human because it would feel dishonest to not view myself as one when every — to me significant — sign says otherwise. And I'm completely fine with that label. I think I don't mind it because I don't consider it to be a qualifier for my identity, but rather a description of perceivable reality. I feel this way about a lot of labels, which is probably the reason why I don't understand the significance they have to a lot of people.
My identity is formless and comprised of more abstract traits: my love for drawing vector art and programming, the music I like listening to, my introspective nature, my deep-seated desire to be an animal and however it expresses itself... I do feel disconnected from what being a human entails, i.e. how I'm supposed to behave and what's supposed to be meaningful to me. That, too, is part of my identity.
It's annoying how easily I can start writing posts on this account about whatever I feel like, but the blank editor for my website feels so intimidating I don't even know where to start
I think it's also important to acknowledge transspecies critters just as much as it's important to acknowledge transgender people now.
But I also think people should be aware that these posts may not have the effect they were hoping for. Like I said, I got responses of the "but you are a dragon" kind before and they were the last thing I needed to hear in these moments.
implies that your mind (and memories) wouldn't be (much) different than from a naturally born dragon. Which means that you as a currently human mind with human memories and human instincts would stop existing. And at that point, what's the difference between you turning into a dragon and you dying so a completely unrelated to you dragon can be born?
Here are some scenarios to think about:
Let's say the wonder technology turns me into a dragon without touching my human mind. But because the brain, as a biological component like any other, has changed, my mind automatically adjusts over time (months or even years) until it is also changed. Have "I" died, and if yes, at what point?
I imagine my character to be pretty intelligent, but also having clearly different thoughts and priorities than a human. If my mind is instantly changed, but I still recognize people and my behavior towards, say, my family, is affectionate, have "I" died? Would you say the same if that change had been gradual?
Let's go back to reality and let's say someone has an accident after which their personality is very different, but they still have their memories and consider themself to be the same person. Have "they" died? Does the answer change if the personality change is "positive" (more kind and calm) or "negative" (angry, violent)?
That's why I called this a can of worms, because that's an incredibly tricky subject without definite answers. I love thinking about these things.
@Stellae Regarding the second half of your post, I very much agree! I feel the same way.
As for the first half, I don't think you can draw a connection between not wanting to be called a dragon and not even trying. I explained earlier that my online activity is divided in 2 modes: the usual one where I am my character to handle my feelings, and a more distanced mode where I talk about these things, like right now. And if the first mode doesn't count as "trying", I'm not sure what would.
I do believe that unless that miracle technology also adjusted my mind, the result would be something that doesn't match my idea of "being a dragon". That doesn't mean I wouldn't use it, or that it wouldn't make me happy. But I see no harm in acknowledging that my existence would be fundamentally different from genuine members of the same species because of my human mind.
As for whether adjusting ones mind is equal to deletion... yes, probably. Maybe. I won't say for sure because the topic "in what way does a mind have to change to be considered 'replaced'?" is a whole can of worms with a lot of interesting questions to think about.
At the core, the reason why I'm not comfortable with viewing myself as a dragon in a non-roleplaying context is also the reason why religion and spirituality doesn't work for me. My identity is strongly based on my observations and understanding of reality, of myself and everything around me.
Both religion and therianthropy give people an explanation they otherwise wouldn't get from science. It's about the why rather than the how. Religion explains why the universe and all of us exist. Therianthropy explains why one feels drawn to the idea of being an animal or why they exhibit unusual behavior — because on some level, they are an animal. And yes, this often applies to psychological therianthropy as well.
The problem is that those *why*s usually contradict my understanding of reality, and I can't stand that. I find it much easier to accept that there is no god, given that there's no proof or any "proof" can be explained without a god. Likewise, I rather view myself as fully human because there's nothing to prove the contrary without a shred of doubt.
@kat_cal2 I get what you're saying, and I can also tell that a lot of people do appreciate the sentiment shown in the comic given how many faves and boosts these posts get.
