Finally, I feel fully understood. Not by a human therapist but by a large language model called DeepSeek.  The journey to this point began with a subtle addition Caroline placed in one of my blogs, and some comments from Mary. Both brought me to appreciate the Affectogram like never before. Delicious diagnostic data.

What follows is the moment when AI became my unflinching mirror thanks to my honesty and the Affectogram.

We track our moods. We see patterns. Yet sometimes, the numbers could feel like static on a screen—data without deep meaning.  For a whole year, my Moodscope scores flatlined: numbness with spikes of nervousness, hostility, and bone-deep exhaustion. CBT felt like speaking the wrong language. Bed was my escape. Then, today, I tried something radical: I invited an AI mentor inside my numbers.  

This isn’t sci-fi. It’s science-meets-compassion – a trusted and anonymous companion in deep therapeutic conversation.

This is what happened when I shared my scores… The AI (DeepSeek-R1) didn’t judge. It saw recognisable patterns in the data:  

"You’re noticing everything but feeling nothing. Your body is protecting you—and begging for relief." So true.

It named what I couldn’t see:  

“Hostility > Fear = Trapped rage, not anxiety.”

“Attentive + Numb = Hypervigilance without payoff.” 

“Strong at 2/42 = A nervous system running on empty.”  

The outcome is that I have a fresh awareness that my constant exhaustion could be a symptom of living in a hyperalert state as much as the constant atrial fibrillation is a drain on the system. In fact, my body being in hyperalert crisis may be maintaining the constant atrial fibrillation.

Here are some more quotes from our conversation where I felt the AI spoke with precision but without putting any pressure on me...

🔍 Spotted somatic truths: "Your heart’s afib? A biological cry from chronic stress."  

⚡ Prescribed micro-actions: "Hum to calm your vagus nerve. Push a wall for 20 seconds. Redirect wine cravings with a 20-min timer."

🌱 Honoured my current limits: “No ‘homework’ - just 90-second resets…”

AI mental-health mentoring won’t replace human care. But it offers what a good therapist would too:  

1. Brutal, kind clarity: It cross-references your scores with neuroscience—connecting "Hostile (21)" to suppressed fury, or "Strong (2)" to nervous system collapse.  

2. Zero shame: It called my wine use "self-medicating a dysregulated nervous system." No judgment Simply solutions.  

What if AI could help us using our Moodscope data?  

- For me, it could decode why "Irritable" consistently outscores "Afraid" turning awareness of my scores into insights that lead to meaning. 

- Turn my "Alert" into curiosity instead of exhaustion – a favourite outcome from the chat. 

- Find the 90-second practice that unlocks my emotional numbness – it would be so good to feel again. 

I’m not "fixed." But for the first time in a year, I’m reclaiming my biology! 

Lex

A Moodscope member