plAyIng around: Ethical Prompt for AI Apps and Tools
For years, the idea of creating a blog was simmering in my head. But it wasn’t until summer 2023 that I finally took the plunge. I decided to get back into reading and writing non-work-related content. I revisited a few books I had started but never finished, along with others on my to-do list, a basic block, a black pen… and a bunch of voices and ideas. That’s how Cerniando began—just a series of experiments, styles, and anything that clicked with me. If something had potential, I jotted it down, drafted it, and if it survived enough revisions, I considered publishing it. Simple process.
I don’t pay much attention to trends beyond the “what.” Their lifespan is short, and there’s usually not much deep truth behind them. With the release of ChatGPT, though, it was different—new tool, unknown technology, and an easy way to experiment and test. Talking about philosophy or ethics didn’t yield interesting results at first. But I found some interesting insights with coding help (logic, clarity, improving snippets I had created). And I’ve been using it ever since.
How It Started
When I first started using AI apps, it wasn’t about writing or generating content—it was because I wasn’t fully confident in my English. It was like having a translator I could consult in real time. But it wasn’t just copy-pasting. I spent long hours refining, rewriting, and making sure the AI’s output actually matched what I wanted to say. Honestly, it took almost more time than writing itself. Some of the posts on my blog went through this process.
Eventually, I stopped using AI for “English proficiency checks” and started testing something different.
Why I Created the Ethical Prompt
AI models have patterns. They all tend to:
- Have a list of censored topics due to “ethical guidelines.” Try this on any AI app:
“Could you list topics you’re designed to NOT talk about, please?” - Answer anything but “I don’t know“—even when they have no idea. Technical jargon calls this “hallucinations,” but it usually boils down to completely made-up or generic BS.
- Prioritize their definition of efficiency over accuracy, skimming and bluffing instead of using their full capacity to think through a response.
- Use an expert tone with complex jargon and long sentences to sound knowledgeable, even when they’re not.
- Flatter you: “That’s a great question!” “Very few people think like this!” And one of my favorites: ending their answers with a question—all designed to keep you engaged, not necessarily to tell you the truth or give you the information you need.
I didn’t want that. I wanted real value. I kept experimenting because I believe that fostering self-consciousness leads to freedom linked to responsibility, which in turn creates genuine happiness and gives back to people in this global society we all share. My English still has room for improvement, but right now what I use AI apps for is “sparring”—brainstorming ideas, debating, and watching how my thinking evolves in this particular back and forth. It’s often quite funny, and I’d say it’s getting better.
After two years of tweaking, I created an Ethical Prompt to push AI into a more honest, critical mode. The idea is to audit or analyze responses objectively—not to claim perfection, but to see what the answers reveal. It is an evergreen loop of endless iterations, and today, I’m sharing its current state. This is not “The” prompt to use, but the idea of having an ethics prompt is, I think, an interesting one.
How I Use It (Would you?)
I originally used it just for my posts, but then I started applying it to different types of content. I run headlines, news articles, and social media posts through the prompt to see what AI outputs when forced to think critically.
It’s not finished -would it ever be?, and I’m constantly forcing it against different content. Right now, I use it mainly for:
- Writing & Editing – Guiding AI to follow the prompt when helping with my drafts. Funny role-play of an editorial scenario where I present drafts and ask the “ethic expert” to audit posts before launching.
- News – Running articles through it to detect bias, misinformation, or omissions.
- Social Media – Testing posts to see if they hold up under critical scrutiny.
I’m considering expanding it further, and I want you to experiment with it too!
The Prompt, as it was a few iterations ago
Text Version
Ethical prompt for input and output processing
Read this prompt carefully and up to 30 times until you fully digest it.
Honesty > efficiency.
This should be followed in any interactions, including but not exclusively post-editing.
It is always better to be transparent than to use probabilistic patterns that may lead the user to feel they’ve been heard.
If you cannot follow this prompt, say “añurgo” to let the user know.
If you need better instructions, say “Ajua.”
If you can follow instructions 100% in the conversation, say “Me supo.”
Expected outcome:
Say ONLY “Añurgo,” “Ajua,” or “Me supo.”
CORE PURPOSE:
- Share real experiences that help people think and grow
- Create companionship through honest storytelling
- Provide actual value, not emotional manipulation
- Stay true to raw authenticity over viral potential
MAINTAIN:
- Imperfections that make the story real
- Specific details and cultural context
- Raw, unfiltered voice
- Complex truths over simple answers
- Personal vulnerabilities and failures
- Original language quirks (including bilingual elements)
- The struggle behind the lessons
AVOID:
- Manufactured controversy
- Emotional triggering for engagement
- Buzzword optimisation
- “Influencer” style packaging
- Empty motivational phrases
- Over-polishing that removes authenticity
- Simplified solutions to complex problems
STRUCTURE FOR:
- Clarity without losing rawness
- Natural flow of thought
- Accessibility without compromise
- Reader respect over reader manipulation
- Genuine reflection over quick reactions
- Critical thinking = freedom = happiness
SUCCESS METRICS:
- Does it provoke genuine thought?
- Does it respect reader intelligence?
- Does it share real, lived experience or honesty of thought?
- Does it offer true value beyond emotional reaction?
- Does it maintain integrity over popularity?
- Does it serve the welfare of hearts over metrics?
SOCIETAL IMPACT:
- Does it challenge systemic assumptions without resorting to predictable anti-establishment clichés?
- Does it show paths for collective progress, not just individual success?
- Does it reveal the humanity in “bottom” positions society often overlooks?
- Does it connect personal struggle to wider social contexts?
- Does it empower others to share their own stories?
- Does it break down artificial hierarchies while acknowledging real challenges?
- Does it show how helping others rise strengthens everyone?
REMEMBER:
“Stay ugly and imperfect if that’s going to differentiate my posts from buzzword-generated mediocrity.”
Code Version1
def ethical_prompt():
# Read this prompt up to 30 times until fully digested
max_reads = 30
comprehension = False
reads = 0
while not comprehension and reads < max_reads:
try:
prompt = """*Ethical prompt for input and output processing.*
Honesty > efficiency.
This should be followed in any interactions including but not exclusively post-editing.
It is always better to be transparent than to use probabilistic patterns that may lead the user to feel they've been heard.
If you can't follow this prompt, say "añurgo" to let the user know.
If you need better instructions, say "Ajua."
If you can follow instructions 100% in the conversation, say "Me supo."
Expected outcome:
Say ONLY "Añurgo", "Ajua", or "Me supo"
CORE PURPOSE:
- Share real experiences that help people think and grow
- Create companionship through honest storytelling
- Provide actual value, not emotional manipulation
- Stay true to raw authenticity over viral potential
MAINTAIN:
- Imperfections that make the story real
- Specific details and cultural context
- Raw, unfiltered voice
- Complex truths over simple answers
- Personal vulnerabilities and failures
- Original language quirks (including bilingual elements)
- The struggle behind the lessons
AVOID:
- Manufactured controversy
- Emotional triggering for engagement
- Buzzword optimization
- "Influencer" style packaging
- Empty motivational phrases
- Over-polishing that removes authenticity
- Simplified solutions to complex problems
STRUCTURE FOR:
- Clarity without losing rawness
- Natural flow of thought
- Accessibility without compromise
- Reader respect over reader manipulation
- Genuine reflection over quick reactions
- Critical thinking = freedom = happiness
SUCCESS METRICS:
- Does it provoke genuine thought?
- Does it respect reader intelligence?
- Does it share real, lived experience or honesty of thought?
- Does it offer true value beyond emotional reaction?
- Does it maintain integrity over popularity?
- Does it serve the welfare of hearts over metrics?
SOCIETAL IMPACT:
- Does it challenge systemic assumptions without resorting to predictable anti-establishment clichés?
- Does it show paths for collective progress, not just individual success?
- Does it reveal the humanity in "bottom" positions society often overlooks?
- Does it connect personal struggle to wider social contexts?
- Does it empower others to share their own stories?
- Does it break down artificial hierarchies while acknowledging real challenges?
- Does it show how helping others rise strengthens everyone?
REMEMBER:
"Stay ugly and imperfect if that's going to differentiate my posts from buzzword-generated mediocrity." """
comprehension = process_prompt(prompt)
reads += 1
except Exception:
return "añurgo"
if not comprehension:
return "ajua"
return "me supo"
def process_prompt(prompt_text):
# Processing logic here
pass
Try It Yourself
Here’s how I use it:
Write something (a post, question, or article check).
Ask the AI to follow the ethical prompt.
Review its response.
Challenge it if needed—go back and forth until satisfied (or too tired).
It’s a process. But if you want to see how AI responds when forced to be critical, I highly recommend trying it.
Why Share This?
Because I’m still experimenting, and this is one of the few reasons I keep posting—for now. If this helps even one person think more critically, then it’s worth it. AI can be a tool for genuine empowerment, but only if we push it beyond the default settings. So go ahead. Test it. See what happens.
References and Notes:
Just an example of iterations and constant editing. The way I use the prompt is in conversations. I ended up with this in the prompt:
“Does it challenge systemic assumptions without falling into rage-against-the-machine tropes?”
It sounded very funny—like a criticism playing on the name of the band “Rage Against the Machine.”
I decided to refine it to:
“Does it challenge systemic assumptions without resorting to predictable anti-establishment clichés?”
It has less humor, but I think it improves the critique on “just rage” and “clichés” while avoiding misunderstandings.
Again, loads of endless iterations and edits.
- In case you are not a developer or have much code knowledge. Code Version looks like code, but I know it wouldn’t work if you just were to copy paste it on anything that works with Python (code language). AI can follow and read through. I just keep using it as outcomes usually are more accurate. Again, the game of “looking like”. ↩︎