Brain tissue that generates its own electricity.

A brain organoid begins as a small population of human stem cells — often reprogrammed from an ordinary blood or skin sample — guided through the same chemical signaling steps that direct a developing brain.

Over the course of several weeks, those cells differentiate into neurons and supporting glial cells, and self-organize into a three-dimensional structure a few millimetres across, roughly the size of a lentil. There is no skull, no body, no senses, no environment to act in. What there is, reliably and reproducibly, is electrical activity.

That activity does not appear all at once, and it does not stay the same. Early cultures fire sparsely, with little structure. Over weeks to months, recordings begin to show coordinated bursts — many neurons firing together, falling silent, then firing again — a signature of maturing connectivity that has been tracked since some of the field's earliest published recordings. An organoid is not thinking in any sense comparable to a brain embedded in a body with a world to respond to. But it is computing something: a living network turning electrical and chemical input into structured, time-varying output.

To see that output, researchers place the organoid onto a microelectrode array (MEA) — a grid of electrodes listening for the electrical signature of nearby firing neurons. The result is a dense stream of spike data: which electrode, at what time, at what amplitude. Turning that stream into something interpretable — firing rate, synchrony between regions, network bursts, large-scale structure — is itself a substantial computational problem, and one with very little standardization between labs.

In 2023, a research group at Johns Hopkins proposed organoid intelligence (OI) as a field in its own right: using brain organoids not only as disease models, but as a substrate for biological computing, testing whether structured information processing — and eventually learning — can be supported by living neural tissue outside a body. Part of the motivation is practical. A human brain runs on roughly twenty watts, a small fraction of what comparable artificial systems require, and biological tissue offers a fundamentally different kind of computational substrate to study and compare against silicon. Part of the motivation is scientific: organoids let researchers ask developmental and computational questions about neural tissue that are difficult, or simply impossible, to ask inside a living human brain.

The field is young. Its first widely cited organoid electrophysiology recordings date to 2019, and as of 2026 there is still no standard analysis pipeline, no agreed file format, and very little publicly available data to work with. Five projects, presented below, are an attempt to work inside that gap — building the tools the field is missing, characterizing what real recordings actually look like over time, identifying exactly where the standard recording method falls short, and naming an ethical question the field has not yet formulated directly.

Reading, shaping, and questioning the signal.

Two more projects were built in this same sequence, aimed at a harder question than any of the three above: not just reading a signal, but shaping one, through reward, penalty, and an organoid's own timing-dependent plasticity. Both stayed in simulation long enough to become working, verified designs — and then both left simulation entirely. What happened when they reached a real culture is its own story, told next.

This is the part where the code stopped talking to a simulation and started talking to a living brain.

September 2026: four designs that had only ever run against synthetic neurons got plugged into two real human neurosphere cultures, on a lab bench in Switzerland, controlled from a laptop, across roughly forty hours of paid sessions. That sentence alone is the headline. A single researcher, no institution behind him, sent code to living brain tissue and got real electrical activity back, on purpose, on command, four separate times. Whatever each individual result turned out to be, that capability now exists and is proven, and everything below it is what got built on top of that proof.

The first thing that capability delivered was a precise, useful correction. The word-length design assumed a longer word would drive more of the tissue to fire, cleanly, the way turning a knob further makes a speaker louder. On real tissue, across three sessions on two cultures, that knob did not turn: same-length words drew different reactions, different-length words drew the same one. That is not a shrug. That is a specific mechanism, ruled out by direct test rather than by guesswork, and every future design in this line now skips it instead of rediscovering the same dead end by hand.

The second thing it delivered was sharper still. The threat-detection design, built on the same expansion-rate signal every animal with eyes reacts to instinctively, worked convincingly in simulation once a genuinely subtle calibration bug was tracked down and fixed. On real tissue, five independently engineered strategies were thrown at it in a single four-hour session, and the sharpest result came out of the direct comparison: wiring more electrodes at once produced a smaller reaction than wiring just one. That single number flips the core assumption the entire design was standing on, and it was caught because the test was actually run, on real cells, instead of trusted on paper.

The third thing was a bet placed in advance, and it already paid off. Before a single real-hardware session was booked for the simplest possible learning task this framework can pose, a rule got written down and locked: one good-looking session proves nothing, only a repeated one counts. That real-hardware session hasn't happened yet, plainly stated. But the rule itself already won a different fight in this same stretch of work: a separate, genuinely promising effect looked completely real across a handful of trials and then dissolved once it was pushed to a proper sample size. The safeguard was tested before its own experiment even started, and it held.

The fourth thing turned a wide-open question into two hard numbers. Does this tissue's own spontaneous activity balance on the same razor's edge between order and chaos that shows up in earthquakes and avalanches of sand? Five and a half hours of continuous listening, thirty-nine thousand readings, checked against that signature from every reasonable angle: no, not on the pooled data. But split the same recording into its two sessions and the picture gets more interesting, not less: the shorter, calmer stretch leaned toward that special balance, the longer, busier one flatly didn't. Both numbers are on the record, side by side, because a real result that raises a sharper question is worth more than a smooth average that hides one.

Four designs went in. Four hard, specific, checkable facts came out, each one now sitting in public with the exact data and code behind it. That is round one of running real experiments on living brain tissue, done alone, done honestly, done to completion. Round two is already being built on exactly what round one just proved works.

References for the work described above, in the order it was discussed:

Two cultures, one platform, one researcher working alone from İzmir, and one standard held the same across every hour of it: nothing gets claimed until it is measured. That standard just cleared its first real test on living tissue. It is about to get a second.

Functional provenance stratification.

The organoid intelligence ethics literature has developed along three largely separate axes: consciousness and moral status, informed consent and the right to withdraw, and the commodification of donated biological material. Each is reasonably well documented on its own. None directly names the following scenario.

This is not purely speculative. Large-scale iPSC biobanking already catalogues donor lines by differentiation efficiency and functional yield — an organoid-intelligence-specific performance axis would extend an existing practice, not invent a new one. The scenario links two risks not previously connected in this combination: retrospective consent-scope violation, when a line is reclassified by a criterion that did not exist at the time of donation, and structural inequity, when sampling bias causes specific donor populations to be disproportionately coded as lower-performing.

I term this practice functional provenance stratification and propose four minimal safeguards: explicit disclosure in consent forms, transparency and auditability of performance metrics, periodic equity analysis across donor populations, and preserved withdrawal rights independent of a line's performance category. This is a conceptual contribution, offered in the spirit of the anticipatory neuroethics already established for questions of organoid consciousness — not an empirical finding.

This work doesn't happen alone.

This project is funded by an Emergent Ventures grant (Mercatus Center, George Mason University) — an unusually trusting kind of support for an independent researcher without institutional backing or a laboratory. Beyond the funding itself, the wider Emergent Ventures community's willingness to engage with early, unfinished, sometimes wrong work has shaped how this project is run in the open — corrections published alongside results, not smoothed over. Thank you to the fund and to that community.

The Whole-Life Dynamics project above exists because FinalSpark granted direct access to real, continuous organoid recordings through their Neuroplatform — data an independent researcher has no other way to obtain. A subscription agreement is now in place for live, closed-loop sessions on that same platform, and it's what makes the work described further down this page possible at all.

Independent, by necessity and by choice.

I work without a laboratory and without formal academic affiliation — an independent scientist using public datasets (DANDI Archive) and direct data access granted by research platforms. Based in Aliağa, İzmir, Turkey.