Depression, epilepsy, chronic pain, and schizophrenia all involve E/I imbalance. If stimulation can selectively suppress overactive circuits, not just activate them, the therapeutic design space expands considerably.
Full breakdown + refs: substack.com/home/post/p-...
We confirmed this with in vivo two-photon calcium imaging in living mouse cortex, watching individual neurons in real time.
This mechanism also explains why sensations evoked by electrical stimulation fade, and points toward stimulation patterns that sustain perception instead.
High-frequency stimulation preferentially drives inhibitory interneurons (parvalbumin+, >100 Hz capable). Excitatory neurons cap out at 15-20 Hz.
Pause at the right moment and inhibitory drive suppresses excitatory output. Stimulation produces net inhibition, not activation.
For ~100 years, neuroscience assumed electrical stimulation excites neurons. Deliver current, cells depolarize, done.
That assumption is wrong.
Jack Szostak, biologist/geneticist (?) and 2009 Nobel laureate along with Elizabeth Blackburn (another immigrant) and Carol Greider (born in California) for helping figure out the crucial roles of telomeres at the ends of chromosomes. He's the author of this powerful piece.
$165,626,742,154.29
That's the cost of the Trump administration's mismanagement, as estimated by a new tool from @ourpublicservice.bsky.social:
federalharmstracker.org/cost-to-our-...
$1.26B lost from terminated NIH grants as of Feb 2026.
$1,028.74 for each person in the country.
Every tenured PI reading this has the same protection.
The question is whether you'll use it.
#ScienceAdvocacy #NIH #SciencePolicy #FirstGen
Bioniclab.substack.com
🗣️ Compassionate conservatism taught me that privilege comes with obligation.
I have tenure. Institutional protection. A platform.
Using those things to defend the system that gave me a chance is not extracurricular.
It is the minimum.
🏛️ The people dismantling the federal research enterprise have never depended on it.
They mock studies with strange-sounding titles because they've never needed the cures those studies produce.
They are counting on scientists being too comfortable, too busy, or too "above politics" to push back.
🎓 I got out through public hs, Pell grants, & NIH, NSF, Univ system that gave a kid from that background a path into science.
They were not abstractions to me.
They were the mechanism.
Now I'm a named Chair tenured full prof. By every measure, the system worked.
💪 That is exactly why I fight for it.
🏦 The IRS froze my bank account in middle and hs. I couldn't buy lunch.
Here's something they don't teach you about that kind of poverty: you let the answering machine screen every call because it might be a debt collector. (B4 caller ID)
After a while, your friends stop calling too.
🛂 I was raised as a compassionate conservative. Core principle: those with privilege owe service to those without.
Ironic. My family had no privilege to speak of.
My parents were green card holders. >$1M in debt. Too afraid to take entitlement programs bc it might cost them their immigration status.
As long as Russell Vought is running the OMB, no one should be confident that funds authorized by Congress will be spent as intended or upheld by the courts. His solitary goal is the consolidation of _all_ spending decision-making in the “unitary executive.”
Hamster is biology...🐹
The path to stable neural interfaces requires engineering informed by cellular mechanism, not material optimization alone.
Hope to see you there.
🔗 cmu.zoom.us/j/9256129562...
BBB breakdown, pericyte loss, astrocytic reactivity, demyelination. Each degrades recording and stimulation in distinct, mechanistically addressable ways.
Clemastine-driven remyelination improves chronic recording. Bulk myelin enhancement does not. Mechanism specificity matters.
The dominant narrative: BCIs fail because of materials and foreign body encapsulation.
Incomplete. The tissue-electrode interface is an actively remodeling microenvironment driven by neurovascular disruption, glial signaling, and metabolic dysfunction.
Next Wednesday, April 15, 1–2PM ET I'm giving a seminar at @CarnegieMellon on why BCIs fail, and why the answer isn't what most people think.
Powerful piece from @safa-science.bsky.social by @jennifer-troyer.bsky.social @jenna-m-norton.bsky.social and Alice Popejoy
scienceandfreedomalliance.substack.com/p/the-cure-f...
The sterile insect technique Knipling invented for screwworms is now being adapted to fight Aedes aegypti, the mosquito carrying dengue and Zika. One "ridiculous" study of fly sex in the 1930s is still generating new public health tools 90 years later. That's basic research.
📢 The FY2027 budget is a proposal. Congress writes the final number. Tell your reps to reject the 19% USDA cut and $602M NIFA research cut. Basic science infrastructure is not a line item to cut. It's what keeps the emergencies from happening. house.gov/representatives
"Eradication is not an event. It's infrastructure." Screwworm was eliminated because the government funded basic research AND maintained the system that research built. It's returning because maintenance was cut. This is the pattern across all federal science right now.
The FY2027 budget proposes cutting USDA by 19% ($4.9B) while screwworm reaches the Texas border. USDA lost 20,000+ staff through DOGE buyouts. FAO monitoring programs were killed. You can't fight a continental pest with emergency declarations and a skeleton crew.
🚨 Screwworm is back. 164,000+ animal cases across Central America and Mexico. Detections 70 mi from Texas. USDA is deploying the same sterile insect technique invented 70 years ago with federal funding. It works. But only if you keep funding the system that maintains it.
🪰 In the 1930s, USDA paid scientists to study the sex life of the screwworm fly. Congress called it waste. Result: a $250K investment that eradicated a flesh-eating parasite from an entire continent, saved ranchers billions, and cut beef prices 5%. Still in use today.
A channel that detects a neuron is not the same as a channel that detects a neuron doing something meaningful. Recording quality needs metrics beyond spike count, ones that capture whether the recorded population retains its network structure.
The BCI field evaluates biocompatibility by asking: did the tissue survive? The better question: does it still compute? Our data show functional connectivity degrades before structural markers raise concern. By the time histology looks bad, the network failed weeks ago.
A neuron that fires is not the same as a neuron that communicates. We tracked functional connectivity between neuron pairs near chronically implanted electrodes. They were alive. Spiking. But uncoupled from their neighbors. The local circuit was fragmented.
