Madame AnxIety — How I Try to Use AI for Self-reflection

The top candidate for my Madame Anxiety (serious look)

How do I play with “AI” tools when I want to talk about my anxieties? Have you tried it yet?

This life hack is not for the faint of heart. As a lifelong warrior with anxiety and neurosis, I try different strategies to score better in an unequal battle. And I’m having fun doing it.

In addition to the various approaches and methods I’ve already proven, I’ve added language models and stable diffusion (in the popular and marketing jargon, “AI” — I loosely call these tools CSH, or the collective stew of humanity) to my arsenal.

Language models, or chatGPT

• These beasts possess one great trait right in their blood; they may be just “word generators,” but they are by far the best at managing to cook up a readable+creative verbal stew preceding the situation we describe;
• It is in their DNA to generate questions when we give them answers;
• And so I sometimes talk to the model about my dreams or anxieties, asking it to suggest reasons why this or that may have dreamed or made me anxious;
• Sometimes, it kicks you out of your thought patterns, helps you start having fun and laughing at things;
• chatGPT is technically sometimes more creative than my therapist here; this is inspiring, but on the other hand, of course, it’s missing the human touch, and that’s a good thing;)
• I think using LLM for therapeutic purposes can be risky for several reasons, so I don’t recommend it yet; still, one can’t help feeling some relief (similar to journaling) after such a “discussion.”

The top candidate for my Madame Anxiety (crazy look)

Stable Diffusion, Midjourney

• The images of Madame Anxiety that you see were generated from prompts that came out of my personal anthropomorphized idea of anxiety as “an outwardly charming, slightly frightened lady capable of convincing a person of impending danger (ideally non-existent), sometimes even a mad lady — in reality, inside a robot entangled with ancient mechanisms set to monitor the slightest threat.”
• Yes, I mostly laughed at that, which was the point of the experiment;
• It raises the question if it is not a great new way of art therapy (or even “arts therapy”);
• I got some stimulating material for further discussion with the therapist.

It is impressive that the underlying common principle behind stable diffusion and anxiety is that their results materialize from “white noise” and “blurred ideas” planted deeply in human experience — both being able to conceptualize into very realistic creations.

So these were dead ends, misunderstood prompts, not exactly my imaginations ;)

How do you imagine your “demons” (I don’t like this term much; it demonizes the issues, but you know what I mean)? And have you ever tried to involve the “collective stew of humanity” in your self-reflection?