Brain & Mind Weekly — Mar 10, 2026

Your brain learns by getting more redundant, not less. Psychopaths don't lack fear — they enjoy it. And a Belgian company received FDA Breakthrough Device designation for a system that bypasses the eye and stimulates visual cortex directly. This week sits at the intersection of old questions getting sharper answers and new tools that make those answers possible: traveling electrical waves that might literally be perception, AI systems that write plain-English descriptions of what individual brain voxels prefer, and a molecular chain reaction that could finally give drug hunters a shared target across different forms of autism.

For decades, the reigning theory was that your brain learns by making each neuron a specialist — maximum information, minimum overlap. A new macaque study says the opposite is happening.

Researchers recorded from visual cortex neurons as monkeys learned a demanding perceptual task over weeks. As the animals improved, their neurons didn't divide labor more cleanly. Instead, multiple neurons started carrying the same information about the same features — like a group chat where everyone repeats the key point. Learning made the code more redundant, not more compressed.

This makes a certain biological sense. If your brain is constantly integrating noisy, uncertain signals (and it is), you want backup copies of important information, not single points of failure. It's the same logic behind why airplanes have redundant flight systems. And it may explain something long puzzling to neurologists: why people can lose significant numbers of neurons and still perform learned skills remarkably well. The zip file theory of the brain was elegant. The group-chat theory might be true.

The Hollywood version of a psychopath is someone who feels nothing when the rest of us would be terrified. A study in Biological Psychology paints a more unsettling picture: the fear machinery works fine. The labeling is what's broken.

German researchers wired up 119 adults with heart-rate monitors and showed them first-person video clips designed to evoke fear, excitement, or nothing much. People scoring high on core psychopathic traits didn't show blunted physiological responses — they showed elevated ones. Racing hearts, the whole package. But when asked how they felt, they described those states with words like "exciting" and "thrilling" rather than "scary."

The body signals danger. The mind tags it as fun. The authors call this the "Fear Enjoyment Hypothesis," and it reframes psychopathy from a simple deficit ("they can't feel fear") into a reward-mislabeling problem ("they feel it and like it"). The sample skewed toward women and non-clinical participants, so generalizing to forensic populations requires caution. But if this holds up across cultures and with brain imaging, it could shift how courts and clinicians think about culpability and intervention — from "they don't know what fear is" to "they know exactly what it is and find it delicious."

Autism is notoriously heterogeneous — hundreds of genetic variants, no single cause. But a molecular study this week found something rare: a potential convergence point.

The chain reaction works like this: excessive nitric oxide (a signaling molecule) chemically modifies a protein called TSC2, tagging it for destruction. TSC2 normally acts as a brake on the mTOR pathway — a master growth-and-metabolism regulator inside cells. Lose the brake, and mTOR goes into overdrive. The researchers found this pattern not just in one genetic subtype but across multiple autism-linked mutations, including SHANK3 variants and idiopathic cases with no known genetic cause.

Here's why this matters practically: nitric oxide synthase inhibitors already exist as research tools. Having a drug-tractable node — a place in the molecular chain where you could intervene with an existing class of compounds — is unusual for a condition this complex. This is still preclinical, and the distance between "plausible molecular target" and "effective therapy" is measured in years and failed trials. But for a field that has struggled to find shared biology across autism's many forms, a common funnel is a big deal.

Most vision-restoration devices require some functioning of the eye's own nerve pathways. Belgian company ReVision Implant is skipping the eye entirely. Their "Occular" system uses a camera on a headset to capture visual information and transmits it directly to an implant on the brain's visual cortex, creating patterns of light points — called phosphenes — that the brain can learn to interpret as shapes and objects.

This week, the FDA granted it Breakthrough Device designation, which accelerates the path to human trials. We're not talking about restored sight in the way you'd imagine — think a low-resolution digital sketch of the world rather than HD vision. But by going straight to the cortex, this approach could help people whose blindness stems from retinal or optic nerve damage, conditions that retinal implants can't touch. ReVision anticipates initial clinical trials later in 2026, with broader studies in 2027. In the larger BCI landscape, China simultaneously elevated brain-computer interfaces to a "core future strategic industry" in its new five-year plan, with more than 10 invasive BCI human trials underway — matching the U.S. count. The global race is on.

If you have treatment-resistant depression, the standard transcranial magnetic stimulation protocol means trekking to a clinic five days a week for six weeks. For someone whose condition makes getting out of bed feel like wading through concrete, that's a cruel irony.

A UCLA study of 175 patients compared the standard schedule (one session daily, five days a week, six weeks) against an accelerated format: five sessions daily for just five consecutive days. Both groups saw significant reductions in depression symptoms on standard rating scales at the end of treatment, with no statistically significant difference between them. Compressing six weeks into five days doesn't just save time — it could dramatically expand access for rural patients, people who can't take extended leave, and those too impaired for sustained outpatient visits. Larger trials are needed, but if the result replicates, this is the kind of logistical breakthrough that matters as much as a pharmacological one.

ReVision Implant's "Occular" cortical vision BCI received FDA Breakthrough Device designation this week, clearing an accelerated regulatory path. The system bypasses damaged eyes entirely, stimulating visual cortex directly via an implanted electrode array paired with a camera headset. Clinical trials expected to begin later in 2026.

GRAB-Ser fluorescent serotonin sensor, developed at Janelia Research Campus and published in Neuron, is now available to labs. The sensor glows in real time when serotonin is released in living brain tissue, enabling researchers to map neuromodulator dynamics during actual behavior — something previously near-impossible at this resolution.

A flexible polymer neural probe from MIT, reported in Science, recorded over 1,000 neurons in freely moving mice for months without scarring. Thinner than a human hair and made from biocompatible polymer, it bends with brain tissue so naturally the immune system barely reacts — a hardware leap that makes long-duration, high-density recordings practical.