AI NEWS PAGECASE FILETOPIC FILE

Biological Computing

RAT BRAIN NEURONS. 5X FASTER VIDEO. 80% CHEAPER.

FILE UPDATED Oct 5, 2026

WHY THIS MATTERS

TBC’s first commercial product is a neuron-derived AI video model announced with AWS on September 22, 2026: built on an open-source text-to-video base, optimized with a proprietary software layer derived from measurements of living neural activity that adds under 0.1% to the model and runs entirely on conventional AI infrastructure.

SOURCES: Amazon Web Services

ZIMBO'S READ ON Biological Computing

Editorial cartoon: a proud AWS engineer in a hard hat unveils a server rack with a rat brain in a glass jar wired into it, while a brown tabby cat watches with professional interest.

THE BRAIN IN THE RACK

THE LATEST

LATEST ON Biological Computing

  1. Oct 5, 2026

    AMAZON IS BRINGING RAT-BRAIN-DERIVED AI TO AWS

    Amazon Web Services

  2. Sep 30, 2026

    THE SOFTWARE LAYER CAME FROM MEASUREMENTS OF LIVING NEURAL ACTIVITY...

    Amazon Web Services

  3. Sep 27, 2026

    CLAIMED RESULT: VIDEO GENERATION 5× FASTER. INFERENCE COST 80% LOWER...

    Amazon Web Services

STILL MORE NOTES

MORE ZIMBO ON Biological Computing

DEEP FILE

THE WHOLE FILE

chronology · facts · older headlines · sources

NEURONS SELL. SILICON JUST RENTS.

The Biological Computing Co. studies living neurons and turns what it learns into software that makes AI models faster and cheaper — its first product, announced with AWS on September 22, 2026, is a neuron-derived text-to-video model claimed to generate 5x faster at 80% lower inference cost. No wetware ships to the customer: a proprietary software layer under 0.1% of the base model runs on ordinary GPUs.

  1. Sep 22, 2026

    AWS PUTS NEURON-DERIVED VIDEO AI ON THE SHELF

    The Biological Computing Co. and AWS announced the world’s first neuron-derived AI video model: a text-to-video model built on an open-source base, enhanced by a lightweight proprietary software layer derived from measurements of living neural activity. TBC claims 80% lower inference costs, 5x faster generation, and better quality than the base model, with the layer adding less than 0.1% to the model and running entirely on conventional AI infrastructure. The plan is to run it on AWS Trainium, deploy via SageMaker AI, and sell through AWS Marketplace; CEO Alex Ksendzovsky called it the first commercial product of TBC’s neuron-discovery process.

    SOURCES: Amazon Web Servicesfirst-party

MORE FILES