physically stabilizing thoughts
On automated reading, engrams, and the chemistry that organizes memory
I have been refining a new workflow these past few weeks, basically an automated Python system I built for collecting and organizing scientific corpuses from PubMed and moving them through notebookLLM. The process extends my reach as a reader and gives me a broader surface for forming ideas. In some ways, the whole platform stabilizes my biological memory into a silicon storage box. It feels like a small upgrade to the way attention can be shaped by the tools around it.
It is late November in SF and I am preparing for a long winter stretch at w.oodland for a focused run of thought work before opening our scientific lab in New York for the next phase of our biotech company. I have been moving between San Francisco, w.oodland, NYC, and Stockholm this year, and the shifts have been stirring the memory systems a little more than usual. Nostalgia is a hell of a drug.
With the workflow tuned and the automation running smoothly, I pointed the whole system at a simple question. What does it take for a nostalgic thought to hold its shape. Modern neuroscience calls the core unit an engram, a physical trace formed by a specific population of neurons that were activated during learning and then biochemically altered so that their coordinated activity can recreate the memory. These identifiable ensembles can be tagged through immediate early gene activation. Immediate early genes are activity-triggered transcription factors that turn on within minutes of neuronal firing and serve as molecular flags for neurons participating in an experience. Remarkably, these flagged ensembles can be reactivated artificially. Optogenetic stimulation of hippocampal engram cells can trigger memory recall without the original cue (PMID:27601979). It remains one of the clearest demonstrations that experience leaves a measurable biological footprint.
The molecular story begins at the synapse. Long-term potentiation is initiated when calcium flows through N-methyl-D-aspartate (NMDA) receptors and sets off a cascade that inserts α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors into the postsynaptic density (PMID:14744210). Spine morphology changes as well. Proteins like Drebrin A stabilize the enlarged structural state of an activated synapse by controlling actin dynamics (PMID:28469009). These modifications are chemical signatures that anchor memory formation.
Allocation provides the engram with its boundaries. Neurons with elevated intrinsic excitability at the time of learning tend to be recruited into the ensemble, creating a competitive landscape that shapes whether memories overlap or remain distinct (PMID:27318935). A few millivolts of difference in firing threshold can decide how two experiences become linked or separated across time.
Transcriptional programs create long-term stability. Gene expression changes move the engram from a transient electrical pattern into a durable architecture. Immediate early genes initiate the path, and sustained transcriptional responses preserve the trace across days or weeks (PMID:32528254).
Engrams distribute across regions. Hippocampal, prefrontal, and amygdala ensembles coordinate during recall, and their contributions shift as memories transition from recent to remote (PMID:32445753). Retrieval is a synchronized state change across these systems, sculpted by the timing and strength of their connectivity.
Neuromodulators tune the engram at every step. Dopamine influences tagging during novelty (PMID:34271016). Stress hormones reshape synaptic strength through effects on calcium and glutamate signaling. Forgetting has its own chemistry, including complement pathways and microglial pruning that modify or erase synaptic connections (PMID:34103208). The shape of a thought shifts with the internal climate.
After reading through all of this, something settled. The deeper we go into the biology of memory, the more the bottom drops out. Each mechanism is distinct and well studied, yet the coherence of the whole system emerges from timing windows, probability distributions, and competition between ensembles. A thought stabilizes inside a landscape that balances precision with uncertainty.
Ideas, memories, choices
all higher-order consequences of matter navigating possibility.
I keep returning to that pier in San Francisco because it carries a specific pattern formed during a moment of transition. The engram still lights when I see the water or feel the cold air before winter sets in. Physiology holds the pattern long enough for meaning to grow around it.
Underneath everything, the chemistry of a thought arises from quantum constraints, ion flow statistics, receptor probabilities, and the physical states available to a neuron at a given moment. We build meaning inside those boundaries. We build memory out of fluctuating energy landscapes. We build our lives on top of the chemistry that allows a thought to hold its shape.
And yeah, remembering is simple, until you try to remember the mechanism, and the memory takes over the explanation, and soon you are spiraling through a kind of cheerful biochemical echo chamber where every ion reminds another ion to remind another ion to remind you that you were trying to remember how remembering remembers the thing you were trying to remember.