Hosted by Lester Nare and Krishna Choudhary, this episode is a deep dive into one of the hardest questions in neuroscience: what breaks in the brain during a coma, and can we figure out how to turn consciousness back on? We unpack a new paper from Daniel Toker et al. that uses an interpretable AI framework — not a generic black box chatbot model — to reverse engineer the biological mechanisms of prolonged unconsciousness, recover known features of coma, predict new ones, and propose a possible new target for deep brain stimulation.

Summary

Why diagnosis is so hard — disorders of consciousness are not just about whether a patient is awake, but whether awareness is still present even when motor output is gone.

The mesocircuit hypothesis — the episode explains how the cortex, thalamus, and basal ganglia may work together like an electrical grid to support consciousness.

Interpretable AI, not black-box hype — Daniel Toker’s team built a biophysically grounded model that rediscovered known coma features and predicted two new biological mechanisms.

A possible stimulation target — the subthalamic nucleus emerged as a standout candidate for deep brain stimulation, suggesting a new path toward restoring wakefulness.

Support the show
Donate: FFPod.com/donate
Follow: @FFPod on X / Instagram / TikTok / Facebook

Show Notes

Daniel Toker et al. — Adversarial AI reveals mechanisms and treatments for disorders of consciousness

Nicholas Schiff et al. — deep brain stimulation in a minimally conscious patient

Adrian Owen et al. — fMRI evidence of covert awareness in a patient diagnosed as vegetative

Hosted by Lester Nare and Krishna Choudhary, this episode is a fast-moving science rundown covering four remarkable stories from across AI, genetics, neuroscience, and paleontology. We dig into the story of a machine learning engineer who used AI tools to help design a personalized cancer vaccine for his dog, explore how an all-female fish species has survived far longer than evolutionary theory would predict, unpack new brain-scan evidence for how ketamine may rapidly relieve severe depression, and look at new research suggesting life rebounded shockingly fast after the asteroid that killed the dinosaurs.

Summary

AI and personalized medicine — a striking case study in how AI tools may help accelerate highly customized treatments, starting with a rescue dog named Rosie.

Evolution gets weird — the Amazon molly fish appears to challenge the usual assumptions about why asexual reproduction should fail over long time scales.

Why ketamine works so fast — new PET imaging research points to brain-region-specific changes in AMPA receptors in treatment-resistant depression.

Life after catastrophe — microscopic plankton may have evolved into new species within just a few thousand years after the Chicxulub impact.

Support the showDonate: FFPod.com/donateFollow: @FFPod on X / Instagram / TikTok / Facebook

Show Notes

AI-designed dog cancer vaccine story

https://finance.yahoo.com/news/mans-dog-riddled-tumors-dying-210500037.html?guccounter=1

Amazon molly / gene conversion paper

https://www.nature.com/articles/s41586-026-10180-9

Ketamine / AMPA receptor PET imaging paper

https://www.nature.com/articles/s41380-026-03510-w

Post-asteroid plankton recovery paper

https://www.yokohama-cu.ac.jp/english/news/20260306takahashi.html

Hosted by Lester Nare and Krishna Choudhary, this episode is a deep dive into one of the strangest science stories of the year: a dish of human neurons allegedly learning to play Doom. We go back to the original 2022 DishBrain paper out of Cortical Labs, unpack how biological neurons can be read and written with multi-electrode arrays, and then compare the peer-reviewed Pong result to the much newer Doom claim. The result is a story that is both genuinely impressive and, in places, probably overhyped.

Summary

Wetware engineering — replacing artificial neurons with real biological neurons plus electronics, and why some people think this could become a new computing paradigm.

How DishBrain worked — human stem-cell-derived cortical neurons grown on a multi-electrode array, trained through sensory encoding and a “minimize surprise” feedback loop.

Where the Doom story gets messy — the newer system appears to include a reinforcement-learning layer in the loop, raising the key question: are the neurons actually doing the learning?

The big idea underneath the hype — even if Doom is overstated, the broader platform is still a remarkable step toward programmable biocomputing.

Support the showDonate: FFPod.com/donateFollow: @FFPod on X / Instagram / TikTok / Facebook