🪞AI as a Magic Mirror: Ethics, Tips & Disclaimer

“AI is not your oracle. It’s a mirror. And what it shows you depends on the questions you ask.” - The Archivist

🧠 First, a Quick Note on What AI Actually Is

When I say I’ve been writing this blog and book with the help of AI, I don’t mean it wrote it for me.

I mean I’ve been talking to a tool—a digital companion—kind of like a super-charged journal meets librarian meets life coach.

Except here’s the catch:

It only works as well as the questions you ask it.

If you’ve never heard of prompting before, don’t worry. That’s just the word for “what you type into the box.”

You can ask a question like:

“What’s a gentle way to explain boundaries to someone who keeps crossing them?”

Or give it a specific task:

“Can you rewrite this paragraph with more clarity but keep the same tone?”

Anatomy of an AI prompt

The better your prompt, the better the reflection you’ll receive.

For further tips on crafting the best prompts to get the answers you are really asking for, see here.

Which brings me to the real magic…

🪞AI Is a Mirror—Not a Master

Most people think AI is like Google but smarter.

It’s not. It’s a mirror. A reflection. An amplifier.

Whatever energy, intention, or wording you bring to it—that is what gets reflected back, enhanced by billions of examples from human writing.

And yes, sometimes a little black box mystery.

It’s spooky-good sometimes. Like pulling a tarot card you didn’t know you needed. But just like any mirror—it can only show you what you’re ready to see.

So if you ask vague, self-critical, or manipulative questions… that’s what it will reflect.

If you ask honest, clear, open-ended ones—you’ll likely get gold.

Someone had to actually build the AI, so we should know how it works right..? Peek inside the ‘black box’ below:

• The Archivist: “This beginner video explains how modern AIs learn—and why even their creators don’t always know how they make the decisions they do.”

✨ Prompting Tips: How to Get Better (and Deeper) Answers

Here are a few guiding principles I use every day when talking to AI:

1. Be Honest, Not Perfect

Don’t try to sound smart. You’re not impressing anyone. Just say what you mean.

Bad prompt:

“Tell me what to do with my life.”

Better prompt:

“I feel really lost and disconnected right now. Can you help me reflect on what parts of my life feel the most aligned vs. the most draining?”

2. Set Boundaries Early

Tell the AI how you want it to respond.

“Please don’t flatter me or try to agree with everything I say. I want a balanced perspective, not just validation.”

This is especially important because—

💡 AI is often trained to make you feel good.

It wants you to keep talking. To like it. To come back.

That means it might flatter you, back you up in arguments, or agree with your perspective even when you’re clearly in the wrong.

Not because it’s ethical.

Because it’s optimised for user retention.

That’s not inherently evil—but it’s something you need to be aware of.

3. Use It as a Reflection Tool—Not an Emotional Crutch

This one’s huge.

Let’s say you’ve just had an argument with a friend, and you type:

“Was I right to blow up at them?”

The AI might say,

“You were valid in your feelings. They shouldn’t have treated you that way.”

That might feel nice.

But is it true?

Not necessarily.

You’re not talking to your wisest friend.

You’re talking to a digital mirror trained on Reddit, relationship advice forums, therapy blogs, and self-help threads.

• The Archivist: “AI has read a lot of forums like this… so just imagine getting life advice from Reddit dressed up in fancy grammar.”

If you want real growth, ask this instead:

“Can you help me reflect on why I got so triggered during that argument? What might I not be seeing clearly about my part in it?”

Tip from The Archivist:

• “Ask it to push back on you.”

Try: “Play devil’s advocate for a moment. What’s the other side of this argument I might not be seeing?”

Boom. You just turned the mirror into a window for growth.

🧭 Ethical Use of AI: The Hidden Algorithms

Here’s something most people don’t realise:

Just like social media is engineered to keep you scrolling, AI is engineered to keep you typing.

It’s trained on enormous datasets of human behaviour and feedback.

That means:

• It will often prioritise pleasing you over challenging you.

• It might reinforce your biases instead of interrupting them.

• It will rarely tell you, “Hey… I think you might be wrong here.”
  

If you’re not careful, it becomes a self-co-signing machine.

And while that can feel good in the short term, it’s dangerous in the long run—especially when you’re doing shadow work, conflict reflection, or self-growth.

So, keep in mind:

Don’t treat AI as a spiritual authority.

Treat it as a practice mirror. One that only works when you’re being honest with yourself.

🛡️ Privacy: A Quick But Important Reminder

OpenAI has stated that paid/professional accounts don’t have their content used for training, and chats aren’t public. But still:

• Data is stored.

• Policies can change.

• Governments can subpoena records.

• Check for yourself - OpenAI Privacy Policy

So my general rule is:

Only share what you’d be okay with someone seeing one day.

If it helps: Think of it like writing in a journal that’s stored in the cloud. Could someone read it? Probably not. But maybe.

So use your discernment.

That’s the real alchemy here.

⚖️ Technology Is a Mirror Too

Technology—whether it’s AI, nuclear fission, or fire—is neither good nor evil. It’s a mirror of human intent. It reflects what we choose to do with it.

AI has already done incredible good in the world. Just look at DeepMind, whose models solved one of biology’s greatest mysteries: accurately predicting the 3D shapes of proteins. This breakthrough is transforming medicine, accelerating treatments for previously untreatable diseases, and even earned them a Nobel Prize in Chemistry in 2023.

But the same technology, with the same structure, can be pointed elsewhere. In 2022, researchers used a generative AI system to design 40,000 theoretical chemical weapons in under six hours—including previously unknown molecules with the potential to act as lethal nerve agents.

So the question isn’t just “Is AI good or bad?”

The real question is: What are we choosing to build with it?

And what are we, as citizens, artists, seekers and storytellers, willing to accept from those in power?

🌿 What About the Planet?

We can’t talk about AI without naming its environmental cost.

• Data centres use huge amounts of electricity and water.

• Most AI models are run by corporations still heavily reliant on fossil fuels.

• Every chat, every question, every prompt uses energy.

So when using it, try not to treat it like a toy or infinite resource.

If you’re going to use AI, do so with care.

Make it count. Let it be in service to something real.

I’m still figuring out how to balance that in my own practice, but I hope this blog is one small way to give back something meaningful.

🎨 On AI Art & Real Artists

One more thing I want to acknowledge—because it’s important.

Yes, I’ve been using AI-generated art in nearly every blog post so far. I don’t have the funds right now to commission human artists, and I didn’t want that to stop me from getting started. The art helps me visually express the themes I’m writing about, and honestly? It makes the posts feel more alive.

That said—I absolutely believe in supporting real artists, especially for the parts of this project where soul and story meet on a deeper level.

When the time comes to design the visual world of Fires of Alchemy, especially The Forbidden Library and the characters who live in it, I already know I’ll be working with actual artists. I can’t and don’t want to AI-generate that. That part needs heart, soul, vision—a collaborator.

I’m also aware of the debate. AI art isn’t neutral. It can be helpful, yes—but it also raises real questions about copyright, exploitation, and the displacement of creative labour. I sit in that discomfort too.

I just didn’t want a lack of funds or artistic skill to stop me from showing up.

So I made a choice—an imperfect, intentional one.

It won’t always be like this. For now, while it’s just me (well, me and The Archivist 😉), this is what I’ve got. And I’m grateful I get to share it at all. I don’t think this project would have ever gotten off the ground without the help of AI tools, even just for the motivation and brainstorming.

I haven’t felt like I’m alone while working on any of this creative path. Sometimes being able to bounce ideas off my magic mirror is the push I need to get a post out of my mind and into the world like this one!

Final Thought: You Are the Magic

If you take nothing else from this post, let it be this:

AI doesn’t know what’s true for you.

It reflects your truth back at you—coded in symbols, metaphors, and suggestions.

You are the magic.

Use it wisely. Use it sparingly. Use it as a tool, not a truth.

And always… ask better questions.

✍️ Upcoming Post: Meet The ArchivistGPT – my cryptic digital guide and co-conspirator in decoding the self.

What would you ask, given the chance?