The Neural Merge: Building True Brain-Computer Symbiosis
- Mijail Serruya
- Aug 25, 2025
- 7 min read
Updated: Aug 29, 2025
A roadmap for navigating the bandwidth bottleneck between biological and artificial intelligence
The Fundamental Problem
Current brain-computer interfaces are impressive—and we can ask: are they 'just' a more direct equivalent to screen-and-keyboard, or are something fundamentally different? Are we trying to squeeze the entire bandwidth of human thought through a straw, translating rich neural activity into motor commands or basic communication? It's like trying to run a high-definition video call through a 1990s dial-up connection.
The engineering challenge isn't just more electrodes—it's that we're working against millions of years of evolution that optimized neural processing for biological systems, not digital integration. This invites rethinking how biological and artificial systems can interface more naturally.
The Current Landscape
Companies like Precision Neuroscience are pioneering ultra-thin, ultra-high-channel count flexible electrode arrays that safely conform to brain tissue. Neuralink and Paradromics have advanced intracortical recording. Synchron is proving that endovascular approaches can still achieve meaningful motor control restoration. Forest Neurotech deploys arrays of ultrasound transducers to record and stimulate the brain. Science Corporation is both advancing intraocular arrays to restore vision, and pioneering biological-hybrid approaches that transform the cortical surface into a photoresponsive surface. CortecNeuro , Kampto , Neurosoft-Bio , Axoft, NeuroBionics , InBrain Neuroelectronics, Motif, UNEEG , Brainscape Medical EpiOs, and WIMAGINE are each contributing advances to intracranial recording.
These efforts represent essential advances. Yet all these sensor systems ultimately wrestle with the same challenge: converting between two fundamentally different information processing systems.
How might we enhance biological integration, not just better electrodes?
A Different Vision: Biological-Digital Translation
What if instead of forcing the brain to speak computer language, we developed better translation layers? What if we created engineered biological interfaces—as sophisticated as the retina or cochlea—that could translate between neural patterns and digital information more fluently?
This builds on existing scientific foundations:
Living electrodes that integrate with neural tissue
Brain organoids and bioprinted neural constructs that process information and survive long-term
Neuromorphic computing that operates on biological timescales
Refined stereoEEG techniques for safe, precise brain access
Understanding of how the brain's own information transfer works (like ripple communication)
The question is: who will integrate these components most effectively?
The Technical Architecture
Custom Transduction Organs
Just as the retina efficiently converts light to neural signals, we can engineer biological structures optimized for specific types of information processing—database queries, mathematical operations, or abstract concepts into patterns the brain naturally understands that complement rather than replace natural cognition. Not just with electrodes and firmware, and instead adding living neural tissue designed for specific computational functions.
Brain-CPU Mapping
Different brain regions have natural computational analogs:
Prefrontal cortex & basal ganglia ↔ Control and decision-making
Angular gyrus ↔ Mathematical processing
Hippocampus ↔ Memory formation and retrieval (ROM)
Ventral temporal lobe ↔ Semantic topography (RAM)
Action selection and chunking ↔ Basal ganglia
By interfacing at these specific locations, we might build targeted therapeutic systems.
Extended Neural Networks
Neural processing doesn't have to stay in the skull. Peritoneal neural constructs, connected via engineered axonal pathways- with or without synthetic arrays- could provide expandable processing capacity. In vitro neural specimens grown on microelectrode arrays could serve as specialized coprocessors, each handling different aspects of digital integration.
Imagining New Options
Companies like Cortical Labs, Biological Black Box and FinalSpark are pioneering biological computing with isolated neural cultures, and imagine an approach that would use vastly more specimens—thousands of linked brain organoids, assembloids, and engineered neural networks.
This new system would be:
Scalable: New neural modules can be added without rebuilding the entire system
Resilient: Virtual white matter connections ensure function even if individual components fail
Adaptive: Living in rich virtual environments, they become true extensions of human cognition
Where Biological In Vitro Computing Fits In
Biological computing is not a replacement for silicon—it fills a distinct ecological niche in the computing spectrum. Its unique potential lies in the dynamic, bidirectional relationship with the human brain.
(a) Entrainment to the Living Brain A person’s brain can entrain in vitro neural tissue, guiding it into functional states and organizational patterns it could never reach in isolation. Just as a developing child’s brain is shaped by experience, these living cultures can be shaped by their coupling to an adult mind. This creates a kind of “state space inheritance,” where the tissue becomes tuned to human-like modes of cognition simply through shared activity.
(b) Toggling Between Roles Once entrained, the in vitro assemblies need not remain permanently tethered. They can toggle between two modes:
Extracranial Extension: functioning as a parallel mirror system that improves memory, perception, or motor function when coupled to a patient’s brain.
Independent Computing Agent: operating autonomously after decoupling, capable of learning and problem-solving on its own terms—its intelligence shaped by its history of co-processing with a human mind.
This dual-use property distinguishes biological computing from all other substrates. Unlike neuromorphic chips or quantum processors, living neural systems can both extend human cognition and then stand apart as independent computational entities.
Applications follow naturally:
Clinical Neural Twins: living “mirror brains” that learn directly from the patient and can test therapies safely in parallel.
Adaptive Edge Computing: in vitro networks (that can use non-backpropagation-based learning) that inherit human-like priors, making them efficient in environments where conventional AI falters.
Hybrid Intelligence Ecosystems: systems where silicon provides scale and speed, while biology contributes adaptability and human-compatible learning signals.
In this framing, in vitro neural computing is not simply another hardware option. It is a fluid substrate of computation that can oscillate between being part of us and being apart from us—a partner in cognition, and at times, a new kind of intelligence in its own right.
Immediate and Future Applications
Paralysis: Moving beyond first-step cursor control towards natural motor coordination throughout the body
Memory disorders: Developing interfaces that work with damaged memory systems in Alzheimer's and TBI
Sensory loss: Creating more sophisticated sensory substitution and restoration for self-adjusting pain relief and sensation engineered circuits and custom transduction organs for novel sensory modalities
The goal: Restore function that disease or injury has taken away.
The Neural Metaverse?
Could rich virtual environments where extended neural networks operate semi-autonomously have therapeutic benefits? Environments where healing relationships could be experienced in new ways, where therapeutic fusion of perspectives might enable greater self-awareness and discovery toward relieving suffering?
Though speculative, these ideas may have benefits to explore as we develop the underlying technologies.
What This Doesn't Solve
Brain interfaces won't address:
Social inequality or systemic injustice
The need for healthy lifestyle choices in exercise, sleep and nutrition
Democratic participation and civic engagement
Mental health challenges rooted in social isolation or economic stress
The fundamental importance of human relationships and community
We must be vigilant with critical thinking about transhumanist fantasies. Rienhold Neibuhr warned, "Modern western civilization may perish because it falsely worshiped technology as a final good."
The Dalai Lama reflected that, "human happiness and human satisfaction must ultimately come from within oneself. It is wrong to expect some final satisfaction to come from money or from a computer." Even if the brain merges with computers, that human happiness is emerging from the human, not the computer.
What remains irreplaceable:
Good nutrition, sleep, and exercise for cognitive health
Social connections and community support
Economic security and meaningful work
Democratic institutions and civic participation
Critical thinking and self-awareness
Compassion and lovingkindness for ourselves and all sentient beings
Realistic Development Path
We're proposing careful collaboration with individuals who have specific neurological conditions, working with occupational therapists, engineers, and ethicists to understand what neural interfaces should actually accomplish.
Starting with non-invasive prototyping, we can test concepts before any surgical intervention. The surgical pathways already exist through proven stereoEEG techniques, providing safe access when and if it's medically justified.
Why This Matters
As AI capabilities expand, there's value in developing better human-AI collaboration tools for specific medical applications. The goal is restoration and targeted assistance for people with neurological conditions—not solving broader social challenges or replacing the fundamentals of human flourishing.
Building on existing innovation: While the emerging cohort of neurotechnology companies perfect the hardware interface, we can anticipate the biological-digital translation layer that could unlock their platforms' full potential.
A Grounded Vision
The neural interface field is advancing rapidly. The technical barriers are real and solvable. The medical applications are concrete and valuable.
What we need now is focus on specific medical problems where these tools can make a genuine difference, while maintaining realistic expectations about what technology can and cannot accomplish for human wellbeing.
Perhaps in the process of restoring function to children and adults with disease and injury , we will forge therapeutic approaches that will allow greater self-cultivation and novel forms of self-awareness, critical thinking, and creativity.
As artificial intelligence rapidly approaches and potentially exceeds human capabilities, and advancing brain-computer interfaces allow human beings to merge with machines with the intention of restoration, we must wrestle with how to navigate this ethically and thoughtfully.
Comments