In 2026, the "heat wave" of brain-computer interfaces (BCI) continues to surge.

 

Recently, Li Xiaojian, founder of Weiling Medical, set a rule: investors who want to meet him can — but at a cost of 5,000 yuan per hour.

 

The reason is not a lack of money, but a lack of time.

 

In this rapidly heating new industry, opportunities for in-depth exchanges with those who truly "understand" are becoming a scarce resource — entrepreneurs, investors, and researchers are all vying to find them.

 

 

Bringing together technology, capital, and industrial forces for exchange was precisely the original intention behind setting up the BCI session at the inaugural "Great Nation New Drugs" Global Conference (CPIC 2026).

 

As an ecosystem platform focused on life science innovation and industrial development, Tongxieyi hoped that this session would allow technological exploration, capital judgment, and industrial demand — which had previously been scattered across different fields — to engage in deep exchange and collision within a single space, jointly exploring the key pathways for BCI to move from technological breakthrough to industrial implementation.

 

During CPIC 2026, Tongxieyi conducted an exclusive interview with Tao Hu, founder, CEO, and Chief Scientist of Neuropixel, engaging in an in-depth dialogue on technology route debates, industrialization path choices, and how Chinese companies can build their own competitiveness.

 

Tao Hu
Founder, CEO & Chief Scientist, Neuropixel

 

Chapter 1: BCI Is More Than Just Medical Value

 

The year 2026 marks a critical juncture for the BCI industry, as the sector transitions from frontier technology exploration to the threshold of industrialization.

 

Policy support has paved the way, providing top-level support for industrial development. Both the 2026 Government Work Report and the "15th Five-Year Plan" have designated BCI as a key future industry track, clarifying its long-term development positioning. Multiple regions — including Guangdong, Shanghai, and Jiangsu — have successively issued supporting implementation measures, establishing official guidance prices for medical services such as BCI-based neuromodulation and brain function rehabilitation, clearing institutional obstacles for clinical implementation and commercial application.

 

Following the policy direction, capital has moved quickly. According to statistics, in the first half of 2026, there were 64 financing events in the BCI field, with total funding reaching 6.68 billion yuan — far exceeding the full-year transaction scale of 2025. Recently, Zhudong Technology announced the completion of a 330-million-yuan angel round, once again setting a new record for early-stage BCI financing in China.

 

The underlying logic behind policy support and capital influx lies in BCI's compound value across multiple sectors — healthcare, technology, and intelligent interaction.

 

The healthcare sector bears the most urgent clinical needs and serves as the core anchor for industrial implementation. Tao Hu pointed out, "The brain is one of the most important and complex organs in the human body. It carries functions such as cognition, movement, and perception, but it is also extremely fragile and susceptible to disease. As a special 'sensor' connecting the brain to external devices, BCI can help us more precisely perceive and identify brain states to assist in disease diagnosis on one hand, and on the other, intervene in abnormal signals through neuromodulation, offering new possibilities for treating brain diseases."

 

It is precisely this unique technical attribute that, in Tao's view, gives BCI its true medical value — not in simply replacing existing technologies, but in solving problems that were previously unsolvable.

 

Within the current clinical medical system, there remain significant technical blind spots for neurological function impairment diseases. Once core physiological functions such as movement, language, and perception are permanently lost due to injury or disease, traditional drugs, surgery, and rehabilitation therapies are largely unable to restore function, leaving many severely ill patients without effective interventions for long periods.

 

Tao Hu summarized these needs as "irreplaceable" — patients with high-level spinal cord injuries cannot move, aphasic patients cannot speak, and blind patients cannot perceive the world.

 

ALS (amyotrophic lateral sclerosis) is an even more extreme example: as the disease progresses, patients progressively lose both movement and language abilities, trapped in a desperate physical and mental state. These are situations that cannot be solved by "improvement" — they require "reconstruction."

 

And this is precisely the significance of BCI exploration. For patients whom traditional medical approaches cannot help, BCI attempts to re-establish connections between people and the external world by capturing and decoding brain signals, offering the possibility of restoring lost motor, language, and other functions.

 

But that's not all. In Tao's view, BCI has attracted attention also because, in the longer term, it carries the future vision of connecting human intelligence with artificial intelligence.

 

"The brain is the most core organ of human beings, and BCI offers a possibility — it enables the integration of the human brain with the increasingly advanced silicon-based intelligence," Tao noted.

 

This also explains why, despite BCI currently being primarily implemented as a medical device, it has been designated by the state as a future strategic industry.

 

In the short term, BCI relies on severe-disease rehabilitation medical care to achieve commercial breakthrough and validate product value. In the long term, it aims to bridge biological intelligence and artificial intelligence through brain-computer interaction — giving the industry immense room for growth.

 

Chapter 2: The Route Debate — No Single Answer

 

As a cutting-edge interdisciplinary life technology, BCI has yet to form a single technological paradigm. The industry is characterized by multiple parallel routes and layered development.

 

Using whether surgical implantation is performed as the core dividing line, BCI can be divided into two major camps: non-invasive and invasive.

 

Non-invasive BCI, represented by devices such as EEG caps, requires no surgical implantation, offering higher safety and convenience. However, because signals must pass through tissues such as the scalp, skull, and meninges, significant attenuation occurs during transmission. They are also susceptible to interference from muscle electrical signals and environmental noise, limiting high-precision control and complex information decoding.

 

Invasive BCI, by contrast, directly captures neural signals through implanted electrodes, obtaining higher-quality data and considered an important direction for achieving high-precision brain-computer interaction. At the same time, the invasive route faces challenges including surgical risks, biocompatibility, long-term stability, and large-scale application.

 

In Tao's view, invasive and non-invasive approaches are not simply substitutes for each other, but rather technological choices corresponding to different application scenarios. "There is no absolute superiority or inferiority among technology routes — the key is whether they match real needs."

 

He further explained that from an application perspective, non-invasive BCI places greater emphasis on safety, convenience, and large-scale adoption, and may be more oriented toward light medical or consumer-grade scenarios such as sleep monitoring, emotional regulation, and attention training. Invasive BCI, on the other hand, targets higher-precision neural signal acquisition — for severe patients such as those with ALS or high-level spinal cord injuries that are difficult to address with traditional technologies, it represents an important exploration direction for achieving complex motor and language function reconstruction.

 

Within the invasive route, the industry also faces different technological choices.

 

Currently, one of the most closely watched companies globally is Neuralink, owned by Elon Musk, which follows a deep-implantation route — using surgical robots to insert hair-thin flexible electrode wires deep into brain tissue to directly capture neural signals. Recently, Neuralink has further introduced a trans-dura implantation approach, improving its surgical robot to complete electrode implantation without resecting the dura mater.

 

Neuropixel has chosen a different path — a flexible cortical electrode route: ultra-thin flexible electrodes only 10 micrometers thick, which attach only to the surface of the cerebral cortex (subdurally), without damaging the brain parenchyma, working by capturing electrical signals from the cortical surface.

 

Regarding the industry-wide question of "whether to benchmark against Neuralink," Tao offered a different perspective: "We should objectively benchmark against the effectiveness and safety of Neuralink's clinical functional restoration, rather than blindly copying its technological path. We insist on an original 'China solution.' The core is not to imitate their appearance or form, but to solve real clinical problems — through flexible cortical electrodes that do not damage the brain parenchyma and a modular architecture, achieving safer, more stable, and more clinically scalable applications."

 

Regarding his own technological route, Tao summarized Neuropixel's approach with three characteristics: no brain damage, no brain heat, and easy scalability.

 

"No brain damage" corresponds to the flexible cortical electrode route. Compared to rigid electrodes that penetrate deep into brain tissue, flexible electrodes can better conform to brain tissue, reducing damage and stability issues during long-term implantation.

 

"No brain heat" focuses on energy management of the implanted device. Neuropixel adopts a modular design that places the battery and processor under the chest skin, reducing the heat burden on the head-implanted device.

 

"Easy scalability" is reflected in the surgical pathway. Rather than relying on specialized robots for implantation, Neuropixel uses the DBS surgical paradigm based on mature neurosurgical experience, lowering the barrier to clinical adoption.

 

As a new industry, the entire sector is continuously iterating and exploring optimal pathways. In the future, BCI holds even more possibilities, and the explorations of various players will drive the industry further.

 

Chapter 3: Showmanship Won't Win

 

As industrial heat steadily rises and the number of entrants continues to grow, how to build core competitiveness and what constitutes the industry's moat have become critical questions for all players.

 

Many people assume the core challenge of BCI lies in electrode materials, chip design, and underlying manufacturing processes.

 

But in Tao's view, the real challenge of BCI is not any single technological point, but how to integrate multiple components into a system that can enter the clinic.

 

Tao explained further that BCI development requires electrodes, chips, algorithms, wireless communication, energy management, and clinical applications to achieve a high degree of synergy.

 

Simply put, electrodes must match chips, chips must match algorithms, algorithms must match communication systems, and ultimately, requirements for implant volume, temperature control, and surgical procedures must also be met. In Tao's words, "showmanship won't win."

 

This is precisely what distinguishes BCI from other technological competitions. Tao noted, "Whether Tesla or SpaceX, the core advantage of Musk's companies has never been the superiority of any single component, but system integration and continuous optimization around the final product goal. BCI is no different."

 

When breakthroughs in individual technologies are insufficient to build a long-term moat, what truly tests a company's strength?

 

According to Tao, the long-term core barriers for companies are concentrated in three dimensions:

 

First, full-chain independent control at the hardware level. From electrodes and chips to operating systems, each link requires years of vertical accumulation. Supply-chain-dependent models cannot build real moats — BCI is a systems engineering project, and system-level optimization requires each component to co-evolve within the same architecture. This kind of architectural capability cannot be achieved by sourcing and assembling.

 

Second, large-scale clinical data and generalizable algorithms. Companies that complete large-scale multi-center clinical trials first will accumulate massive real-world, high-quality EEG data and validated generalizable algorithm models. This is a first-mover advantage granted by the time window, which latecomers cannot simply replicate through catching up.

 

Third, ecosystem connectivity. Seamless integration with various external devices determines whether the technology can move from medical assistance to life empowerment. Whoever can build the most extensive device-connection ecosystem will transform first-mover advantage into a sustainable moat.

 

"The stacking of these three barriers makes it very difficult for latecomers to replicate simply by burning money," Tao said.

 

— Tongxieyi Summary —

 

BCI is still in its early exploratory stage, but it is opening a new window connecting life sciences and artificial intelligence.

 

From helping severely ill patients restore motor and language functions, to exploring the possibility of integrating human intelligence with AI, the value of BCI continues to expand. But for such an interdisciplinary, long-cycle future industry, true breakthroughs depend not only on single-point technological innovations, but also on long-term synergy among research, industry, and capital.

 

This is the value of CPIC 2026's focus on BCI — bringing innovative forces from different fields together in a shared space for exchange and collision, promoting further connections among technology, industry, and application needs.

 

How far BCI can go in the future depends not only on how technological boundaries are pushed, but also on how the industrial ecosystem is collectively built.