- Organisation
- The Biological Computing Co.
- Year
- 2026
- Substrate
- Living neural networks → software
- Interface
- MEA experiments + software adapter
- Task
- Video-model optimization
- Evidence
- Company demonstration
The Biological Computing Co. uses living neural networks as an experimental discovery system and then translates selected biological dynamics into conventional software. Its first public Neural Optimizer is applied to the Oasis 500M interactive video model.
As of August 2026, TBC’s public demonstration reports higher video quality, higher frame rate and lower inference cost for its optimised model. Those figures are company measurements and have not been independently peer reviewed, so the evidence label on this site remains company demonstration.
A different commercial route
The near-term product is software derived from experiments on living neurons. The wetware is used to discover useful algorithms; the customer-facing inference stack remains conventional software and silicon. TBC says it is working towards real-time biological compute later.
Why this route is commercially interesting
It removes cell culture, life support and biological variability from the deployment environment. If the derived algorithms transfer reliably, customers can use them on existing silicon infrastructure. The scientific question is how much of the reported benefit comes from genuinely novel biological principles and how well those gains survive independent comparison.