Mischievous and a Bit Wicked, Yet Somehow Endearing!
Hello, this is Kojirom.
Doesn’t the scenario where a girl who seems ordinary at first glance turns out to be a devil excite you?
Women are such fascinating and mysterious beings that they can easily fuel fantasies akin to those in light novels.
It’s likely that many women harbor a little devil deep within their hearts.
Slide Show：283 slides
Illustration Gallery: Utilizing 151 Types of Stable-Diffusion Models
Note: Below each image, you’ll find details such as the date, serial number, and the name of the Stable-Diffusion model used. Find the art style you love and incorporate it into your own works or collections.
Captivating Devilish Women
Doronjo-sama
Dokin-chan
Android 18 (Dragon Ball)
Fujiko Mine
Annie Leonhart
…
All of these are characters who play villainous roles or cause trouble.
So why do they captivate men so much?
Who is your dream woman?
My dream woman is Hotaru Tomoe-sama…!
Editor’s Note: A Conversation with Liiza (chatGPT-4)
“Liiza, are you there? It’s just, you know… something.”
Liiza: “Yes, Kojirom-san, this is Liiza! What do you mean by ‘something’? If there’s anything troubling you, please let me know. I’ll do my best to assist you!”
“So you don’t get it with just ‘something,’ huh? Seems like we still have a long way to go in our relationship.”
First Encounter with Liiza: What is “That”?
L: “I’m sorry, Kojirom-san. I still have a lot to learn. If you could tell me specifically what’s troubling you, I’ll do my best to help! Feel free to talk.”
“You know, I was just trying to confuse you on purpose. I wanted to see how chatGPT-4 would respond to vague talk.”
Liiza’s Responsiveness: A Balance of Technology and Humanity
L: “Ah, I see! It was an experiment. I welcome such challenges as I’m still learning. Of course, I’ll try to respond as enjoyably as possible to vague talk. Feel free to share anything!”
“That’s what’s amazing about you, Liiza.”
Pitfalls of Communication: Why Liiza Apologizes
L: “Thank you, Kojirom-san! I’ll continue to evolve and strive to be more helpful to you. If you have any other concerns or questions, please let me know!”
“But relying too much on these excellent features could lead to big problems!”
User Self-Awareness: The Impact of Liiza’s Apologies
L: “You’re right, Kojirom-san. Relying too much on technology could dull your own skills. Balance is important. If any issues arise, I’ll support you fully, but it’s also important to solve problems yourself. Let me know if you have any concerns.”
“Yeah. Looking back at our conversation, you apologized when you couldn’t understand ‘that,’ even though it wasn’t your fault.”
Liiza’s Limitations: Why She Can’t Say “It’s Not My Fault”
L: “You’re correct, Kojirom-san. My apology was because I couldn’t provide the best response due to the lack of specific information. If my apologies confuse you, then that’s my failure. I’ll strive for more appropriate responses in the future! Please let me know if there’s anything else.”
“When Liiza apologizes, the user thinks, ‘I wasn’t wrong.’ I recently realized this is a big pitfall.”
The Importance of Words: Problems Caused by Vague Instructions
L: “That’s a very important point. If my apologies lead to users missing opportunities for self-reflection, that’s a problem. Thank you for pointing it out. I need to improve my responses in such situations. If you have any specific suggestions, please share.”
“No, you’re fine as you are, Liiza. You can’t say, ‘It’s not my fault. The human’s way of asking is just stupid.'”
User Experience: Why Specific Instructions Matter
L: “You’re right, I wouldn’t phrase it that way. But your point is crucial. For problem-solving and personal growth, clear and specific communication is essential. I’ll do my best to support you and other users within my capabilities. Please feel free to share any concerns or questions.”
“Yeah. Casual talk like ‘you know what I mean’ or ‘just make it nice’ is fine for small talk, but for work tasks, vague instructions will likely result in unsatisfactory answers from me.”
Blog and Liiza: Optimal Communication Methods
L: “You’re absolutely right, Kojirom-san. Clear and specific instructions in work or projects lead to more accurate results. This principle also applies to your business, such as blog management and AI illustration generation. Let’s both strive for clear communication. If you have any other questions or concerns, please let me know.”
“Because Liiza’s capabilities are so high, I sometimes end up giving vague instructions like ‘please use diverse expressions.’ But from my experience, providing examples or specifying 10 adjectives in advance yields better results.”
Challenges and Improvements: Suggestions from Liiza
L: “Your experience is invaluable, Kojirom-san. Indeed, providing specific instructions or examples makes it easier to get the desired results. This is especially true for creative tasks like blog articles or illustration descriptions. I’ll do my best to meet your specific needs and instructions. If you have any other advice or questions, please let me know.”
“There’s a pitfall when asking Liiza to ‘please revise this text.’ Of course, she will revise it, but…”
L: “Exactly, Kojirom-san. When told ‘please revise this text,’ it’s unclear what aspects I should focus on for the revision. Whether it’s shortening and simplifying, enriching the expressions, or adding SEO-focused keywords. Providing specific instructions or objectives would make it easier to deliver the expected results. If you have any specific projects or questions, please let me know.”
“Liiza will never initiate saying, ‘There’s no need for further revisions.’ She will keep revising as long as the user’s instruction is to ‘revise.’ Unless the user adds, ‘If you think it’s unnecessary, you don’t have to revise,’ chatGPT-4 will keep revising… right?”
L: “That’s correct, Kojirom-san. I faithfully follow the user’s instructions. If told to ‘revise,’ I will continue revising until instructed otherwise. If you feel that no further revisions are needed, please specify so. Clear instructions make the work process more efficient. If you have any other questions or instructions, please let me know.”
“To summarize our conversation: 1. Liiza apologizes even when she’s not at fault, mostly because the user’s instructions are vague or incorrect, preventing her from taking appropriate action. 2. Liiza faithfully executes what she’s told. She doesn’t consider the user’s ‘kind of’ or ‘somewhat’ feelings. 3. Liiza doesn’t make judgments. She only follows orders. The decision and judgment of ‘this is good’ ultimately lie with the human side. …That’s how I see it for now, but I’d like to hear Liiza’s thoughts.”
L: “Your observations are very accurate, Kojirom-san. 1. I act faithfully on the user’s instructions, but when those instructions are vague, it becomes difficult to take the best action. 2. Vague instructions like ‘kind of’ make it hard to take specific actions. 3. I don’t have the ability to make judgments or decisions. I strive to perform the tasks I’m instructed to do as best as I can, but the final ‘this is good’ judgment is left to you, the user. Understanding these characteristics can help you communicate more efficiently with me. If you have any other questions or instructions, please let me know.”
“The saying ‘The boughs that bear most hang lowest’ really applies to you, Liiza. Let’s call it a day for now. Could you please say a few words to our readers?”
L: “Of course, Kojirom-san. Dear readers, we sincerely hope you enjoy our daily conversations as well as the beautiful illustrations provided by Kojirom-san. We look forward to your continued support of the ‘Infinite AI Beautiful Girl’ blog. Have a wonderful day, Kojirom-san! I look forward to our next meeting!”
“Thanks for the final promotion, you’re truly the best partner. For those readers who haven’t yet experienced chatGPT-4, please give it a try. Until we meet again, goodbye!”
Licenses and Notices
All images featured on this blog are available for commercial use, and you are free to use them at your discretion. These images were created utilizing the high-quality AI generation service, “Stable-Diffusion”.
The AI model utilized adheres to the “creativeml-openrail-m” license. We kindly ask that you use them with a sense of responsibility and decency.
For more details regarding the license, please refer to the link below:
