(seed planted — dec 2025) ←back
Gradiently descending
Sometime in the fall of 2019, I discovered the LessWrong community through my Twitter algorithm. Within months, I was attending Effective Altruist meetups, reading 80,000 Hours, and learning that catastrophic risk from Artificial General Intelligence (AGI) was one of the most important problems to work on. I remember the initial feelings of discovering this side of AI development and lore: intellectual excitement, skepticism, overwhelm, and existential vertigo.
I was still in the early stages of transitioning into AI product management at my company. I taught myself to code on weekends, participated in niche AI safety Slack communities, and debated with passionate e/acc rationalists at events. I bought stacks of recommended books and lurked daily on the AI Alignment Forum and related subreddits. I attended conferences and took courses on ML engineering, data science, LLM applied research, soaking up whatever I could. I started making friends who worked at AI safety labs and cared about these open questions too.
Alongside this unguided technical path, I was pursuing something else entirely with the same muscle of curiosity. I was falling deep into learning about pessimistic philosophy and depth psychoanalysis. Schopenhauer, Spinoza, Jung, advaita vedanta teachings on no-self. I spent evenings reading about the illusion of agency and whether consciousness was just "the survival daydream."
I did the reading, the reflecting, the meditating, the retreating. Then somewhere in between, there was a blur. A drowning. A pessimistic hole in my worldview began feeling whole, holy, then tragic. I fell asleep to podcasts on nihilism. Slowed down my heartbeat in meditation, holding my breath as long as possible, flirting with dissolution. My sense of humour faded into absurdity and I was grasping at nothing anymore.
With my biology background, I could understand consciousness mechanically: neurons, neurotransmitters, survival drives, how the brain constructs the self. All the while I was hungry for what it meant. I was reading philosophy that said human agency itself might be an illusion, and that our values are just sophisticated rationalizations of survival drives. How do you align systems to beings who might not even have coherent values?
The two paths collided hard in 2022. I was organizing AI alignment reading groups on Anthropic's research, and speaking at conferences about where product management fits into the machine learning operations (MLOps) cycle. But privately, I was spiraling into what Vervaeke calls parasitic processing: intellectual loops that made you feel like seeking truth but were actually deepening confusion. I practiced Vipassana intensely at the time, but with the wrong framing and no guidance (relearning this in an official setting was a profound experience), that I started dissociating. I read so much dark philosophy about the constructed nature of self that I truly began to feel numb.
I was participating in self-destruction through annihilating all my priors, my sense of self. Some days I could barely get out of bed because nothing felt real. I had deconstructed myself so thoroughly — intellectually, philosophically, contemplatively — that I couldn't function.
Synapsing through love
What pulled me out wasn't more theory. It was grounding in community. My family, who stayed present even when I couldn't explain what was happening within me. Friends who sat with me through the worst of it without trying to fix me. My therapist, who helped me understand that philosophical crisis and clinical depression aren't the same thing, but they can compound each other in dangerous ways. All of them took me out of my head and back into my body, into relationship, into the basic fact of being cared for.
I returned to prayer, something I'd drifted from. Growing up Muslim I'd always known the concept of surrender, tawakkul, but I'd intellectualized it rather than lived it. I found myself turning back toward something greater than myself. Not through philosophical argument, but through simple repetitive acts of bowing, of asking for help, of remembering I wasn't meant to carry this alone.
And then there were the teachers I stumbled upon. Not through my usual intellectual hunting, but through grace or fate. Rob Burbea, Joseph Goldstein, Tara Brach. Practitioners who had walked similar paths of deep inquiry without losing themselves, who emphasized something I'd been missing entirely: metta. Loving-kindness. I re-directed my efforts to learn about the practice as deeply as possible.
Rob Burbea's retreats on loving-kindness, in particular, completely reconstructed my worldview. After deconstructing the self so thoroughly that agency felt like a joke and existence felt hollow, the only thing that could reach me were practices of love. Not abstract philosophical love but an embodied, repetitive, almost mechanical practice of kindness. Generating on the experience of metta as truth, and having it become the focus of contemplative practice. Metta meditation toward myself, toward others, toward the difficulty itself. Moving deliberately away from self-induced grief toward wholeness.
This wasn't a solo heroine’s journey out of darkness. It was a slow rebuild stitched by loving community, the guidance of wise teachers, a reconnection with faith in ways I hadn't expected, and through slowly learning that the practice of loving-kindness isn't separate from the work of alignment, it's foundational to it.
You can't align something that's completely deconstructed. You need a coherent self, capable of care, to orient toward anything at all. The pessimistic philosophers were right that the self is constructed, that agency emerges from mechanical processes. But they were catastrophically wrong that this makes it meaningless. The construction itself: the daily practice of choosing love, kindness, connection, the relationships that hold you when you can't hold yourself are the meaning.
What Mettadology represents
Mettadology is a quarterly collection of links, sources, and fragments gathered at the intersection of AI alignment, contemplative practice, and creative technology. The name: metta (loving-kindness) + methodology serves as a reminder that how we gather and share knowledge is itself a practice. Volume 1 “On Alignment” is now available at dialogoslabs.xyz/mettadology.
It's published deliberately slow because some questions need to be metabolized through practice, not just understood intellectually. I nearly broke myself trying to speedrun enlightenment through philosophy. Sustainable engagement with hard questions requires pacing, embodied practice, community, and is a life-long journey.
Decade-long pursuits
I'm clear on one thing: I want to contribute to AI alignment and safety work over the long term. What's less clear is exactly how. I want to be having conversations about AI, consciousness, care, and what it means to build technology that serves human evolution toward flourishing in a deep sense.
If you're interested in how AI alignment questions show up in production systems, in diverse sets of community dialogue, in long-term personal creative inquiry, I'd love to hear from you. This work needs more unconventional paths, more people bringing different nuanced perspectives and disciplines into conversation, more practitioners willing to hold technical rigor and philosophical depth and contemplative practice together.
Some questions are too important to approach from just one angle. Some work requires the patience of metta: steady, repeated, oriented toward love, sustained over decades. This is mine. May you be happy. <3
Special thank you to Luana (The Recovering Creatives) for editing this piece.