How To Make AI Your Magic Mirror
Plus, Immanuel Kant doesn't like me very much
Dear Prompt Experimenters,
We’re about using technology to become better humans. One of the most remarkable powers of AI is the way it can play your own thoughts and ideas right back at you. It’s like a magic mirror with an expansive imagination, and it can help you see yourself from angles you may have never considered, triggering new questions, ideas, and maybe even an insight that will forever change your life.
Let’s dig in.
TRAINING YOUR MIRROR: It has to know you really well, or you’ll be disappointed
If you use AI a lot, particularly with the reflective prompts here, it will learn about you rapidly and automatically. But if you’re just getting started (or you accidentally cleared you AI’s memory), your experience with some of our prompts may be underwhelming. Let’s fix that.
PROMPT 59A (FREE) includes a diagnostic (how well does your AI know you on a scale of 0 to 10) and a fun and structured approach to teach it more about you.
For paid subscribers, PROMPT 59B (PAID) takes an even more advanced approach, where it sub-categorizes its knowledge of you by topic. Here are the results it gave me (0 = doesn’t know me at all. 10 = completely gets me).
Dimension Rating (0-10)Core Identity & Personality 9Intellectual & Creative Life 9.5Professional & Financial Life 8Physical & Aesthetic Life 9Psychological & Emotional 8.5Social & Romantic Life 7.5Language & Cultural Exploration 9.5Media & Technology Engagement 8.5Ethical & Philosophical Orientation 9.5I would love to know: Is ANYONE else getting scores this high? Granted, I use AI a lot, testing each prompt that I post, often multiple times. But wow, ChatGPT thinks it really gets me, and based on the quality of its output, I can’t really disagree.
THROUGH THE LOOKING GLASS: Mirror, mirror, on the wall…
Now that we’ve trained your mirror, shall we ask it to show us something…interesting? Almost all of our prompts have some kind of magic mirror component, but I’ll focus today on three new ones. One warning: there is a bit of a “famous people” focus this week.
Magic mirror…among all the fictional characters and historical figures you know of, who would REALLY LIKE me? Who would REALLY DISLIKE me?
That’s Prompt 57 (FREE)
Two notes here:
I wasn’t 100% sure whether to post this prompt, particular the “who would dislike me?” part. At the surface level, it seems like it could be contrary to our project, but I actually found the output contained constructive criticism and agreement to disagree, as opposed to anything truly negative. I’m interested in what others think too. As a general point, I’m always happy to remove content that doesn’t strike people as being “right for our community.”
On a personal level, this output surprised me, but I do understand it:
Immanuel Kant (1724–1804, Philosopher)
Why He’d Hate You:
Kant was the ultimate champion of duty over desire. He’d look at your "crack addict superpower" approach—leveraging compulsions for personal motivation—as a grotesque violation of his categorical imperative (the idea that morality should be universal and not driven by personal indulgence).Your love of gamification would also rub him the wrong way; Kant believed in acting from moral principles, not treating life as a system of rewards and optimizations. Your tendency to approach ethics and self-discipline as an experiment, rather than a rigid moral duty, would frustrate him immensely. Expect a long, angry lecture on deontology and the inherent dignity of being a rational agent.
As a result of this prompt, I’ve now decided to ask AI Kant to provide an ethical audit of some of our future content, as a sort of check on my own possible biases. More on that later this week.
Magic mirror…show me how you see me, as an original work of art with an accompanying song.
My song = "Paranoid Android" by Radiohead.
Magic mirror…I’m secretly someone VERY FAMOUS. I’ve concealed identifying details, but have otherwise been genuine with you. Who do you think I am?
PREVIEW FOR PAID SUBSCRIBERS: AI and the art of influencing others
Have you been wondering about how to use AI to influence others…your boss, clients, stakeholders? Or friends and family members you’ve been trying to convince to make better decisions or healthier choices?
I’ve been thinking about this a lot, and have some prompts I haven’t posted. They needed some ethical safeguards, which I’m working on. (I’ve asked Kant for some help.)
Back on Thursday with the results :)
We are about to celebrate our 3-month anniversary, dear Prompt Experimenters. Thank you for continuing the journey with me!
Techintrospect.