For nearly eight decades, computing has followed the same fundamental rule: intelligence runs on silicon.
Every major breakthrough from personal computers and smartphones to cloud computing and generative AI has been built on increasingly powerful semiconductor hardware.
Cortical Labs is testing a radically different idea.
Instead of making computers behave more like brains, the company is exploring whether actual living neurons can become part of a computing system.
That idea recently captured public attention when Cortical Labs demonstrated brain cells learning to play the video game Doom using its CL1 biological computing platform.
The gaming demonstration generated headlines. The patents reveal something much bigger.
Taken together, Cortical Labs’ patents suggest the company is not building a novelty brain-powered computer. It is developing the infrastructure required to make biological computing practical, trainable, and scalable.
And if that vision succeeds, the implications extend far beyond gaming.
What Is CL1?
CL1 is a biological computing platform that combines living human neurons with electronic hardware.
The system allows neurons to receive information from a digital environment, respond to stimuli, and adapt their behavior through feedback.
In simple terms, the neurons are not merely being observed.
They are actively participating in computational tasks.
The recent Doom demonstration showcased how these biological neural networks can learn from interactions with a digital environment and improve performance over time.
But Doom is not the destination.
It is simply a controlled environment for measuring learning.
The patents provide a clearer picture of where Cortical Labs appears to be heading.
The First Challenge: How Do You Turn Living Cells into Reliable Hardware?
Patent: US12223409B2 – Biological Computing Platform
The first hurdle in biological computing is surprisingly basic.
Living neurons are not transistors.
They require carefully controlled environments, continuous maintenance, and stable interfaces with electronic systems.
Patent US12223409B2 focuses on the architecture needed to integrate biological neural networks into a computational platform while maintaining their functionality.
At first glance, it may appear to be a patent about keeping cells alive.
Its strategic significance is much larger.
Without a reliable biological substrate, biological computing remains a laboratory experiment.
This patent addresses the foundational infrastructure layer required to transform living neurons into a usable computational resource.
What This Reveals
The patent suggests Cortical Labs understands that commercialization begins with stability.
Before biological computing can compete with traditional hardware, it must operate predictably and consistently.
Just as data centers require reliable hardware infrastructure, biological computing requires reliable biological infrastructure.
This patent represents an early attempt to build that foundation.
The Second Challenge: How Do You Teach Neurons to Perform Useful Tasks?
Patent: US20230133430A1 – System and Method for Training In Vitro Neurons
Traditional AI systems learn through software training methods.
Living neurons do not.
They must be trained through environmental interaction and feedback.
Patent US20230133430A1 focuses on methods for training biological neural networks using structured stimulation and response mechanisms.
The patent describes systems that allow neurons to receive information, generate responses, and adapt based on feedback.
What This Reveals
This patent highlights a critical insight.
Growing neurons is not the difficult part.
Teaching them consistently may be.
The value of biological computing does not come from having neurons in a dish. It comes from turning those neurons into systems capable of performing useful computational work.
The Doom demonstration should be viewed through this lens.
The game is not the innovation.
The training methodology is.
Just as machine learning transformed software through training frameworks, biological computing may depend on the development of repeatable neuron-training architectures.
The Third Challenge: How Do Computers Communicate with Neurons?
Patent: US20230134609A1 – Hybrid Optical and Electrical Training System
Even if neurons can survive and learn, another challenge remains.
Computers and neurons speak fundamentally different languages.
Digital systems communicate using electronic signals.
Neurons communicate through biological and electrochemical activity.
Patent US20230134609A1 focuses on hybrid optical and electrical stimulation systems that allow information to flow between biological and digital environments.
The patent introduces mechanisms for delivering stimuli and capturing responses using multiple communication methods.
What This Reveals
This patent may be the most strategically important of the three.
Every computing platform requires an interface layer.
The keyboard became the interface for personal computing.
Touchscreens became the interface for smartphones.
Application programming interfaces (APIs) became the interface for software ecosystems.
Biological computing requires its own interface layer.
This patent suggests Cortical Labs is investing heavily in solving that problem.
Without communication, neurons are biology.
With communication, neurons become part of a computing system.
The Bigger Picture: Cortical Labs Is Building a Computing Stack
Viewed individually, the patents appear to solve separate technical challenges.
Viewed together, they reveal a coherent platform strategy.
The portfolio addresses three foundational requirements:
- Maintaining biological neural hardware
- Training biological neural networks
- Creating communication pathways between neurons and machines
These are not application-specific inventions.
They are enabling technologies.
That distinction matters because enabling technologies often create more long-term value than the first applications built on top of them.
The patents suggest Cortical Labs is attempting to establish intellectual property around the infrastructure layer of biological computing rather than a single end-use product.
The Doom demonstration attracts attention.
The patents reveal the underlying platform strategy.
What Happens If Biological Computing Becomes Commercially Viable?
The immediate question is not whether biological computers will replace traditional computers.
That is unlikely in the foreseeable future.
The more important question is where biological systems could outperform conventional architectures.
Artificial Intelligence
Modern AI models require enormous computational resources and energy consumption.
Biological neural networks evolved to process information using remarkably little energy.
If biological computing can be scaled, it could introduce alternative approaches to adaptive learning and pattern recognition.
Robotics
Robots operating in unpredictable environments often struggle with adaptation.
Biological neural systems may provide advantages in learning from changing environments without extensive retraining.
Drug Discovery and Neuroscience
The same platforms used for biological computation could become valuable tools for studying neurological disorders, testing treatments, and understanding learning mechanisms.
Defense and Autonomous Systems
Organizations seeking adaptive computing systems with lower power requirements may find biological computing increasingly attractive for specialized applications.
Edge Computing
One of the biggest constraints in edge AI is energy consumption.
Biological computing introduces the possibility of highly adaptive systems operating with dramatically different power requirements than current AI hardware.
Whether that promise becomes reality remains uncertain.
What is clear is that biological computing creates a new category of competition rather than simply improving existing computing models.
Who Should Be Paying Attention?
The obvious answer is AI companies.
The more interesting answer is everyone whose business depends on the future of computing.
Semiconductor Companies
NVIDIA, AMD, Intel, TSMC, Samsung
The threat is not immediate displacement.
The risk is that biological computing opens a parallel path toward computational intelligence that does not depend entirely on transistor scaling.
For decades, performance improvements have come from better silicon.
Biological computing challenges that assumption.
Neuromorphic Computing Companies
BrainChip, SynSense, Intel’s Loihi Program
These companies attempt to mimic biological neural systems using silicon architectures.
Cortical Labs is exploring the alternative approach using biological neural systems directly.
The two paths may eventually compete for similar markets.
Brain-Computer Interface Companies
Neuralink, Synchron, Precision Neuroscience, Blackrock Neurotech
These companies focus on connecting computers to living brains.
Cortical Labs is bringing living neural systems into computing environments.
The technological overlap around neural interfaces, stimulation, signal interpretation, and feedback systems could become increasingly important.
Artificial Intelligence Leaders
OpenAI, Google DeepMind, Anthropic, Meta
The current AI race focuses on larger models, larger datasets, and larger computational infrastructure.
Biological computing introduces a different question:
Can intelligence be created more efficiently rather than simply at greater scale?
That question alone makes biological computing worth monitoring.
Pharmaceutical and Neurotechnology Companies
Johnson & Johnson, Roche, Novartis, Eli Lilly, AbbVie
Platforms capable of training and monitoring living neural networks may eventually create opportunities in neurological research, drug discovery, and disease modeling.
For these companies, biological computing is not only a computing story.
It is also a biology story.
What Should Companies Do Now?
For IP Teams
Track continuation filings and future patent families around biological computing, neuron training systems, and neural-machine interfaces.
One patent is an invention.
A growing cluster is a strategy.
For R&D Teams
Focus on the bottlenecks identified by the portfolio:
- Biological stability
- Training methodologies
- Interface architectures
These appear to be the areas where long-term differentiation may emerge.
For Competitive Intelligence Teams
Monitor patent activity alongside scientific publications, partnerships, funding rounds, and product announcements.
The strongest signals will emerge when multiple indicators begin pointing in the same direction.
For Business Leaders
Do not evaluate CL1 as a gaming demonstration.
Evaluate it as an early signal that alternative computing architectures are moving closer to commercialization.
The timeline may be uncertain.
The direction is becoming increasingly visible.
The Strategic Question Is Not Whether Brain Cells Can Play Doom
The Doom demonstration is fascinating because it challenges our assumptions about learning and intelligence.
But the patents point toward a more consequential question.
Can living neurons become a practical computational resource?
Cortical Labs appears to believe they can.
Its patent portfolio is not focused on a game.
It is focused on the infrastructure required to make biological computing possible.
If biological computing remains a scientific curiosity, these patents will represent an ambitious experiment.
If the technology matures, they may represent some of the earliest building blocks of an entirely new computing architecture.
And that possibility is far more significant than a high score in Doom.
This analysis is based on publicly available patent publications and information related to Cortical Labs’ CL1 biological computing platform available at the time of writing.
As biological computing continues to evolve, new patent filings, continuation applications, partnerships, funding activity, and competing technologies may significantly reshape the landscape.
Want to track the latest innovation signals in biological computing?
Request the updated patent intelligence analysis to access recent patent activity, emerging competitors, technology developments, and strategic opportunities shaping the future of biological computing.