████████╗██╗███╗   ███╗
 ╚══██╔══╝██║████╗ ████║
    ██║   ██║██╔████╔██║
    ██║   ██║██║╚██╔╝██║
    ██║   ██║██║ ╚═╝ ██║
    ╚═╝   ╚═╝╚═╝     ╚═╝
 ██╗      █████╗ ███╗   ██╗████████╗██╗███╗   ██╗
 ██║     ██╔══██╗████╗  ██║╚══██╔══╝██║████╗  ██║
 ██║     ███████║██╔██╗ ██║   ██║   ██║██╔██╗ ██║
 ██║     ██╔══██║██║╚██╗██║   ██║   ██║██║╚██╗██║
 ███████╗██║  ██║██║ ╚████║   ██║   ██║██║ ╚████║
 ╚══════╝╚═╝  ╚═╝╚═╝  ╚═══╝   ╚═╝   ╚═╝╚═╝  ╚═══╝
 

> PhD candidate, biomedical engineering @ Columbia University

> invasive modulation for augmented neural control

> author @ Palladium Magazine

> NYC

Pen 1: true hand pathPen 2: Kalman decodePen 3: Ridge decodeDecoders trained on active reaches, tested two ways. Hover on the left and the plotter draws eight centre-out reaches; both decodes stay on the hand path. Move right and the test trials become passive bumps: the Ridge map, with no model of how a limb moves, curls away from the path, while the Kalman filter's dynamics prior keeps its trace close. The readout is hand-velocity R² computed from the plotted traces, set to the measured values (Area2_Bump, monkey Han).
The sending agent and its messageColony: cage cards · Ops: booking drum · Data: chart recorderThree agents share one end-to-end encrypted group chat. Colony keeps the breeding cards, Ops books the equipment, Data makes the figures. Every message is sealed at the instrument that sends it and opened only at the one it is for. The relay in the middle forwards it and lights a lamp for each channel, but its padlock never opens: it routes what it cannot read. Move across the bench to choose who speaks.
Flow-focusing junction, where oil pinches the aqueous stream into dropletsThree syringe pumps on lead screws: nuclei (back), gel beads (middle), partitioning oil (front)Nuclei loaded at λ per droplet; pointer x sets λHow single nuclei end up in their own droplets: the pumps push nuclei and barcoded gel beads into one stream, and oil from both sides cuts it into droplets at the junction. Nuclei arrive at random, so the count per droplet is Poisson: with mean λ, a fraction 1 − e^−λ of droplets hold a nucleus and of those, (1 − e^−λ − λe^−λ)/(1 − e^−λ) hold two or more. The readout's doublet figure is that fraction of filled droplets. Loading low keeps doublets rare at the cost of mostly empty droplets. The rate assumes 16 µL/min of aqueous into 65 pL droplets; the drawing runs far slower.
The filter cube in the light path; pointer x turns the turretDAPI · ex 358 / em 461 nm: every nucleus in the sectionTH (Alexa 488) · ex 495 / em 519 nm: noradrenergic neurons of the locus coeruleustdTomato · ex 554 / em 581 nm: the Cre reporter in the same neuronsHow three stains are read from one section: each cube pairs an excitation filter, a dichroic mirror and an emission filter, so the lamp excites one fluorophore and only its light reaches the eyepieces. DAPI lights every nucleus across the slice; TH and tdTomato light the two small bilateral clusters of the locus coeruleus beside the fourth ventricle. Overlaying the TH and tdTomato images is how the histology pipeline scores colocalization.
syllogism.txt

i. "Society is founded not on the ideals but on the nature of man, and the constitution of man rewrites the constitutions of states."

— Will & Ariel Durant


ii. The nature and constitution of man are subject to change via intelligent design through biomedical engineering.


iii. We can improve societies and the constitutions of states via intelligent design through biomedical engineering.