In March 2026, Science Corporation successfully placed the world’s first biohybrid brain-computer interface in a human patient at Yale New Haven Hospital. The device, developed by Max Hodak—Neuralink’s co-founder and former president who left in 2021—combines lab-grown neurons with electronics to read brain activity. Unlike Neuralink’s approach of inserting electrodes directly into brain tissue, Science Corp’s pea-sized sensor sits on the brain’s surface and uses hundreds of thousands of living neurons to form biological connections. The implant successfully recorded neural signals, validating a radically different path in the race to build brain-computer interfaces.
Neuralink’s Ex-President Bets on Biology Over Silicon
Max Hodak co-founded Neuralink in 2016, left in 2021 amid long-delayed human trials, and quietly founded Science Corporation the same month. Five years later, he’s placed a biohybrid brain implant in a human using an entirely different approach than his former company.
Neuralink uses 1,024 ultra-thin electrode threads inserted directly into brain tissue via surgical robot. It’s proven—21 patients enrolled as of February 2026, with participants playing video games and controlling robotic arms using only their thoughts. The trade-off? Tissue damage and potential signal degradation from scarring over time.
Science Corp’s device sits subdural—inside the skull but on the brain’s surface. It contains 520 recording electrodes plus hundreds of thousands of lab-grown neurons that grow downward, forming synaptic connections with the patient’s brain tissue. Hodak’s quote captures the philosophy: “We’re not just recording from the brain — we’re becoming part of it.”
The technical difference matters. Penetrating electrodes damage tissue—thousands of neurons die per probe insertion. Biohybrid neurons form biological connections naturally, potentially avoiding the foreign body encapsulation problem that plagues all implanted electronics. Theoretically, 1 million embedded neurons could establish 1 billion synapses with brain tissue, vastly exceeding Neuralink’s current thousands of channels.
This isn’t just another BCI startup. It’s the former president of Neuralink proving there’s an alternative path. If biological integration works at scale, it could solve the longevity problems of current BCIs. If it doesn’t, Neuralink’s penetrating approach remains the gold standard.
Living Neurons as Biological Bridges
Science Corp’s biohybrid technology sounds like science fiction: lab-grown neurons embedded in electronics, genetically modified to respond to light, forming living connections with your brain. However, it’s grounded in established neuroscience.
The device embeds stem cell-derived neurons in a thin-film electrode array. These neurons are genetically modified with light-sensitive proteins—a technique called optogenetics that’s been used in research for years. When placed on the cortical surface, the living neurons grow downward, forming synaptic connections with the patient’s neural circuits.
Communication works both ways. Hundreds of thousands of micro-LEDs on the chip stimulate the neurons on demand, sending signals into the brain. Recording electrodes detect when embedded neurons fire in response to brain activity, reading signals back out. The first human implant successfully demonstrated this—neural signals were recorded through the living neuronal layer, validating years of preclinical work in animal models.
The promise is compelling: biology’s 1 billion synapses versus electronics’ thousands of channels. Nevertheless, the question is practical viability. Living cells need nutrients, can die, and may trigger immune responses. Manufacturing devices with living neurons consistently across patients is a challenge no other BCI company has tackled. This is where theoretical bandwidth meets biological reality.
Related: NVIDIA Ising: Open-Source AI Solves Quantum Computing Bottlenecks
1 Patient vs. 21, But $1.5B Valuation Shows Confidence
Science Corp has 1 human patient versus Neuralink’s 21. That gap is real. Neuralink’s participants have logged over 2,000 cumulative days and 15,000 hours of usage. One ALS patient narrated and edited a YouTube video using only brain signals. Science Corp’s single implant has hours of data, not months.
But momentum tells a different story. Science Corp raised $230 million in Series C funding in March 2026 at a $1.5 billion valuation—just weeks before the first human implant. Total funding now stands at $490 million since 2021. Investors include IQT, the intelligence community’s strategic investment arm, suggesting applications beyond medical use: communication in hostile environments, cognitive enhancement, classified defense capabilities.
Dr. Murat Günel, Chair of Yale Medical School’s Department of Neurosurgery, joined as scientific adviser to lead U.S. trials. The clinical trial strategy is pragmatic: target patients already needing brain surgery—stroke victims requiring cranium removal, tumor resections—piggybacking on necessary procedures to minimize additional risk.
Timeline expectations are tempered. Dr. Günel says full trials beginning in 2027 would be “optimistic.” That’s the right framing. BCI trials move slower than companies predict. Consequently, Science Corp isn’t trying to beat Neuralink on patient count today—they’re trying to prove a fundamentally superior approach over the next decade.
Is Biological Integration Brilliant or Too Complex?
The central question: Does adding living cells to electronics solve the foreign body problem, or does it add fatal complexity?
The biological integration case is strong. Traditional BCIs degrade over time. The body encapsulates foreign objects, signal quality drops, electrode impedance rises. Biohybrid neurons theoretically improve over time as synaptic connections strengthen. They avoid tissue damage from penetrating electrodes. They could achieve bandwidth orders of magnitude higher than current systems.
The skepticism is equally valid. What happens if embedded neurons die? Can they be replaced, or does the device become inert? How consistent is neuron-brain integration across different patients? Manufacturing challenges scale exponentially when your device contains living cells. Regulatory hurdles will mount as Science Corp moves beyond early feasibility studies—the “no significant risk” argument that avoids FDA approval initially won’t hold at scale.
Neuralink has FDA Breakthrough Device designation. Science Corp is bypassing that path for now. That’s faster short-term but potentially problematic long-term.
Hodak left Neuralink because he believed there was a better way. This is his shot to prove it. One patient with successful neural recordings is encouraging. But Neuralink’s 21 patients, thousands of hours of usage, and demonstrable functionality—controlling robotic arms, playing games, sending messages—represent proven technology. Science Corp needs to close that gap while validating that biological complexity is manageable, not insurmountable.
Key Takeaways
- Science Corp placed the first biohybrid brain-computer interface in a human in March 2026, using lab-grown neurons to form biological bridges with brain tissue.
- Max Hodak (Neuralink’s former president) is racing to prove a less-invasive approach works, betting on biological integration over penetrating electrodes.
- The technical difference is fundamental: Neuralink inserts electrodes into tissue (proven, 21 patients), Science Corp sits on the brain surface with living neurons (1 patient, theoretical advantages).
- Momentum is building: $1.5B valuation, $230M Series C, Yale partnership, intelligence community investment—but Dr. Günel calls 2027 trials “optimistic.”
- The big question remains unanswered: Does biology solve the foreign body problem, or does it add unmanageable complexity? Long-term human data will decide.
