Memories, dreams, more AI stuff

Day before yesterday, I went to a comic store I haven't went to in years and got two volumes of Bocchi The Rock. There was Dandadan, which I've been trying to read. I'm at volume 7 and I also have volumes 8 and 9. But while the store had volumes 11 and 12, they didn't have volume 10. I also went to a good restaurant and ate flan for the first time. Very good. And common around here, actually. The whole experience was great; but unfortunate that not all of the family could come.

Yesterday I dreamt my account got hacked by an account called "Are You Even Human?", which somehow got access to the account, and while I successfully removed their access to the account, I also deleted my own channel in the process.

Then I dreamt about HUNTR/X, the Ghostbusters, and the Dandadan characters (dubbed "Team Supernatural" in the dream - Okarun, Momo, Aira, Jin, and I think even Kinta) being in the same class. I watched Half Life VR But The AI is Sentient all the way through recently, so I wonder if the series had influence on this dream, or if Team Science will be there in the future.

I'm just gonna spout out nonsense about AI

• AI agentic browsers. Perplexity's Comet. ChatGPT Atlas. Opera Neon. Gemini in Chrome. Copilot Mode in Edge. Dia, made by the people who abandoned Arc because they thought Dia was cooler. They honestly do seem like toys. Privacy used to be the buzzword of technology. Now it's AI, and these browsers don't care. They want to add things to your shopping list. All of them are Chromium. I wanna say Comet and Atlas are Mac-exclusive.
• Microsoft Mico. New AI mascot. Probably does the same things as Copilot. You can turn it into Clippy. Clippy would never do this.
• ChatGPT released Sora 2. That won't be a nightmare. For some reason it has its own "social media" filled with other Sora videos. Sora videos are watermarked by default, but you can bypass that. People are making racist videos of famous dead people. 

Getting into philosophy

This is going to be bare-bones but I...

I use AI. Privately. I used it occasionally in the past for public works (it's the reason Shion Igarashi is a former delinquent don't tell anyone), and my series Pearl Princess Pretty Cure and Pretty Cure Smile Star were AI-assisted, however, both series have since gotten rid of their AI roots. Or at least Pearl Princess did. Well, okay, the name Tama Hamasaki was AI-generated and she's only blonde because of AI BUT the name was just too good sorry probably not going to change it in spite of Tamaki Hamasaki.

There are plenty of reasons why AI is bad:

• LLM AI's sample data comes from the internet, with the majority of it coming from copyrighted works (If you create something, it automatically gets copyrighted. You have to turn it public domain, creative commons, etc., yourself, not the other way around).
• Companies are using AI in spades and is affecting jobs, particularly creative jobs and factory jobs.
• The spread of AI-generated misinformation or "hallucinations" caused by the very nature of AI.
• Inappropriate use of AI in regards to real faces, particularly of women and children (Perhaps I should use "humans" or "all people" instead of "real faces", especially that "fake faces" harm real people too, but I don't know, "real faces" seems right for what I'm insinuating).
• AI slop, scams, and spam. Scams and spam pretty much all use AI in the modern day, while slop can be more easily produced, flooding the media market of choice for capital gain.
• Over-reliance on AI can make you dumb. Also blindly believing in AI can make you dumb but even if everything AI says is true, it can make you think of generating things instead of actually doing them.
• Environmental impact. Though it's a bit messy, though I heard some things.
• AI uses a lot of electricity, so it needs to use a lot of water to cool its systems.
• I heard some people say some of the water AI uses "disappears", but I need to be smarter in order to believe that sometimes water can't just evaporate into the sky to be used another day. But it's probably true.
• It is said that AI uses 10x the amount of energy as a regular Google search. Though nowadays Google searches might take 10x more energy than it used to /j 
• Unfortunately, a lot of AI impact to the environment is undisclosed. 

But the philosophical reasons... well, actually, they don't matter in terms of more immediate, less metaphysical harms. But philosophy and ethics are what underlie why we do things. So I have questions. For example:

• Does AI do everything for you? Do "prompts" matter? Does "direction" matter?
• Are you using AI to help you or to ease you? Like making things quick and easy for you... or even using it as a dopamine boost because "wow I am not typing this myself this is so cool".
• Where does support for AI start and end? Because it seems like a lot of people just love love every single piece of technology that was made up to 2014, minus cryptocurrency (which isn't exactly AI but it is technology), and that Sophia robot. And they love that one AI that detects cancer cells before humans can. But many say that Generative AI should stay out of the arts. Yet Kasane Teto, which uses neural GenAI, is loved. Some projects that use GenAI, like Glorb, who sings and writes lyrics themself but uses AI-generated SpongeBob character filters, and Neuro-sama, who uses various sorts of GenAI technology, are loved, especially since they have a lot of human effort.
• The word "soulless" gets thrown around in AI discussion, though I think is more important is that the word "soul" gets thrown around in AI discussion. One may say "I don't believe in souls, is AI okay now?" The response would be "you don't have to believe in souls to not support AI for the 'there are plenty of reasons why AI is bad' reasons listed above." Okay, then why are you throwing around phrases like "there are souls in all human art" like it's candy if it's so irrelevant then? 
• Many anti-AI sentiments say to support humans instead of AI. Technically not objective but definitely fair and just. But what does that mean? And how do we go around it?
• And many anti-AI statements say "it's doing NOTHING. it is not HUMAN. it has no SOUL. you need FLESH AND BLOOD. reject artificiality." That's where I'm probably most hung up about, which I'll get into sometime, probably not in this blog post, or even on this blog, but some time. 

I've also been trying to figure out why I use AI:

• The world is not easy. It's not going to hold your hand. AI companies purport their product as an assistive tool first and foremost. As someone who felt like I couldn't do anything, I would probably find solace in AI.
• AI can generate things quickly. In seconds. The country is in a "convenience culture". I think a YouTube comment said that a single Simpsons defines American culture, where Skinner's mom wanted all of her groceries in one bag whilst demanding the bag not be heavy. I usually ask for "niche" questions or generations. Instead of asking for answers on a forum, or even drawing or writing or doing the thing I message AI to do myself, I go the easy way out.
• Additionally, Google is worse now. While I have extensions that make the experience better, Google has a lot of AI. I did manage to hide the Google Gemini AI from searches, though! There's also an extension that removes certain sites from Google search, namely sites whose sole purpose is for AI. But even then, the sites Google shows were forced to appeal to its SEO. 
• Validation. The "yesman" praise that is very common in AI. I heard someone say it's a common feature in all LLM chatbots, but I do not know if it's intentional or just a consequence of its programming. Like when AI says "your message is more profound than 99% of the people who chat with me." But the "yesman" praise of AI is a way of AI saying "I'm here and I'm listening to you".
• Novelty. It's a brand new toy. It can do all sorts of things. ChatGPT actually just had a "dev talk" where they introduced a new AI agent tool, and added some third-party apps users can "directly talk to" on ChatGPT, including Spotify and Zillow. It's like when a gacha game adds a new character or a toy line adds a new vehicle and outfits for all of the characters.

And my own limits. I used to be less strict, but now my personal rules are: no videos, no music, no pictures except if you ask really nicely with two cherries on top, use Microsoft David/Zira/Mark instead of AI TTS, limit AI speech synthesis to those Google AI podcasts and even then you should not do that, no AI templates, no AI coding, and only use AI creative writing if you're willing to completely throw it away and forget about it, meaning, do NOT post it publicly on any circumstance.

As for AI research... a lot of people do permit it as long as human-written references and resources are involved, though for some, it's a no-go. One Redditor, for example, said "asking if using AI is cheating is pro-AI propaganda; AI is a plagiarism machine and that's why it should never be used."

I've been using it a lot for philosophy, and what I trick myself into believing isn't psuedo-therapy but it totally is. My philosophy teacher told me not to use AI, not just because of misinformation, but also because of misuse of terms. Philosophy uses many common words that take new meanings in philosophy, something AI would either misunderstand or attribute both meanings to when the common meaning may not apply at all to something. And using LLMs for therapy is bad is because AI is not human.

And I'm finally going more away towards AI because if predictive LLMs, by their very nature, are not good for the epistemological, they cannot be good for the metaphysical.

In my own "AI generations", too, AI has messed up. Once it said that [group of male characters of human intelligence 1] recognizing and respecting the sexual boundaries of [group of female characters of human intelligence 2] is scarier than if [group of male characters of human intelligence 1] didn't. Hello? Do you know what "Man vs. Bear" is even about?
Another time it called a manipulative milkman character's tactics "soft white power" (rather than "soft power"). You can see the problem with that.

I realized I can't transition neatly to my next topic, so I'll just go out and say it.

One popular post against AI is, "Why would I look at this if you didn't have the time to make it yourself?". It's about people having the dignity to respect audiences by showing them something they actually made instead of something they prompted.

But I've been thinking, does this apply to AI-generated content not made to see the light of day? "Why would I generate this if I didn't have the time to make it myself?"

So I've been researching dignity. And I found this article on the Stanford Encyclopedia of Philosophy and while I read the whole thing, I got disappointed that I was actually thinking of integrity instead of dignity.

I haven't read this article yet but I think it will help me on AI.

Yeah. That was two hours of research for nothing.

Also I had a dream about The Amazing Digital Circus episode 7 and for some reason Kinger tortures Jax. And Pomni and Jax go to the supermarket and that helps Jax become a better person. And then Caine starts dancing with pink guards and Precure Kigurumi dancers to "Koi wa Chaos no Shimobe Nari" from the Nyaruko anime and "UFO" by f5ve.