From Data Intelligence to Biological Intelligence: An Intelligent Revolution of Human-Machine Collaboration

Human exploration of "intelligence" is crossing the boundaries of traditional disciplines. Particularly between computational science and life science, the originally relatively independent research paths are moving toward deeper integration. 

This trend is clearly reflected in the scientific breakthroughs recognized by the Nobel Prizes. In 2024, the Nobel Prize in Physics focused on Artificial Neural Networks and Machine Learning, while the Nobel Prize in Chemistry honored computational protein design and structure prediction; in 2026, the Nobel Prize in Physiology or Medicine turned its attention to optogenetics — a technology that enables scientists to precisely regulate the activity of specific neurons through light. 

These seemingly distinct breakthroughs present two mutually reinforcing research threads: on one hand, AI draws inspiration from living systems and in turn becomes a critical tool for exploring life mechanisms; on the other hand, advances in neuroscience are constantly deepening human understanding of the biological basis of intelligence. 

Along this path of two-way evolution, an early-stage cutting-edge interdisciplinary direction is emerging — bio-intelligence. 

Behind its rise lies the practical demand for the self-development of AI. In recent years, AI represented by large models has developed rapidly, but problems such as rising difficulty in obtaining high-quality data and sharp increase in computing energy consumption have become increasingly prominent. Although embodied intelligence has promoted AI to move from the digital space to the physical world, there is still room for improvement in its autonomous learning, continuous adaptation and flexible decision-making capabilities in complex and changing real scenarios. In contrast, the capabilities demonstrated by living systems in energy efficiency, learning and environmental adaptation provide important inspiration for exploring new intelligence paradigms. 

As brain-inspired computing draws on neural mechanisms, brain-computer interfaces establish human-machine channels, and embodied intelligence enables machines to move into the physical space. Along these directions, how will bio-intelligence collaborate with data intelligence and physical intelligence to support higher-level intelligence? 

Technological Shift: Why Does AI Start to Seek Answers from Biological Systems?

The current popularity of AI has long gone beyond the technology circle. The capabilities of large models continue to leap forward, robots are accelerating out of laboratories, and capital is constantly raising expectations for the next generation of intelligent technologies. 

Beneath the boom, the expansion of AI capabilities is unfolding along a deeper path: from data intelligence to physical intelligence, and then to bio-intelligence. 

In layman's terms, data intelligence enables AI to "understand data", represented by large models that learn patterns and generate content from massive amounts of information; physical intelligence enables AI to "walk into reality", with embodied intelligence as an important carrier, perceiving the environment and taking actions through sensors and physical devices; bio-intelligence enables AI to "learn from life", trying to draw on the operating mechanisms of biological systems such as the brain to find computing and interaction methods with lower energy consumption and higher efficiency. 

The capital market provides practical evidence for this evolution path. At present, OpenAI is seeking its latest round of financing, with a pre-money valuation set at 1.25 trillion US dollars; in October this year, embodied intelligence enterprise Figure is in negotiations with NVIDIA for an additional investment of about 1 billion US dollars. 

In the field of bio-intelligence, early-stage capital has begun to make layouts. Australian company Cortical Labs, which focuses on combining lab-cultured human brain cells with computer chips, raised 10 million US dollars in a 2023 financing round led by Horizons Ventures; US company The Biological Computing Co. (TBC) focusing on applied biological computing completed a 25 million US dollar seed round of financing in 2026; US company Bios Life, which focuses on the direction of biological digital twins, completed a 25 million US dollar seed round of financing in 2026. It uses genomic, clinical and other data to build continuously updated digital twins of patients for cancer screening and monitoring. 

These financings reflect different industrial stages. Large models have entered a period of heavy capital and large-scale investment, and embodied intelligence has begun to raise funds for mass production and scenario delivery; bio-intelligence is still in the early stage of the technology curve, the industry pattern has not yet taken shape, and capital flows more to technology verification and engineering. 

However, the market growth potential of relevant sub-tracks has begun to emerge. The Business Research Company, a global market research and consulting firm, predicted in its 2026 report that the global neuromorphic computing market for artificial intelligence will grow from 2.04 billion US dollars in 2025 to 9.28 billion US dollars in 2030, expanding by about 4.5 times in five years. 

Neuromorphic computing is only one of the many technical directions of bio-intelligence. From biological digital twins, brain-inspired chips, brain-computer interfaces to living cell computing, organ chips, and bio-intelligent agents, bio-intelligence is exploring different industrialization paths, and its potential applications extend from low-power intelligent terminals to disease research, drug development, new human-computer interaction and other fields. 

In one sentence: Bio-intelligence is still a blue ocean market at present, but it has high growth elasticity. Once living computing, brain-inspired chips and brain-computer interfaces cross the product verification threshold, bio-intelligence is expected to move from early exploration to the industrial acceleration period, form differentiated breakthroughs and phased development potential in some sub-scenarios, and evolve in parallel and integrate complementarily with the two tracks of data intelligence and embodied intelligence. 

Apart from the market space, the energy consumption constraint of the traditional computing path is another reason why bio-intelligence has attracted attention. 

"Why is data-driven intelligence alone not long-term sustainable? We can calculate the energy account of all mankind." Professor Miao Feng, Dean of the School of Advanced Science and the Institute of Bio-Intelligence of Nanjing University, told 36Kr, "At present, the computing volume of AI training and reasoning is exploding exponentially. If we add up the global energy consumption spent on data movement, it is even equivalent to the power generation of several Three Gorges Dams. In the future, the computing power demand of large models will jump by several orders of magnitude." 

In contrast, even including wind power, photovoltaic power and nuclear power, the total amount of energy available to humans still relies heavily on fossil fuels, and the overall global energy supply maintains a slow linear growth. 

The energy consumption dilemma is rooted in the "separation of storage and computing" of the von Neumann architecture: frequent data movement between memory and processor generates huge energy consumption and latency. 

In comparison, the human brain does not need to encode the real world into "0" and "1". After millions of years of evolution, the human brain demonstrates excellent and efficient perception, processing, learning and reasoning capabilities with extremely low power consumption. 

In fact, research related to bio-intelligence has a long history. In the early stage, it mostly focused on understanding the generation mechanism of human or animal intelligence from the levels of DNA, cells and behavior. Zheng Hairong, Member of the Standing Committee of the Party Committee, Executive Vice President of Nanjing University, and Academician of the Chinese Academy of Sciences, combined long-term research and practice to give a new interpretation of this concept, and included human-machine integration into the research scope of bio-intelligence. He regards bio-intelligence as the core pillar of superintelligence, and puts forward original concepts such as "brain-to-brain communication". 

If described in a more understandable path, bio-intelligence is moving from "understanding life" to "integrating with life": first study how the brain and living organisms generate cognition, then draw on neural mechanisms to develop new technologies such as brain-inspired computing and brain-inspired vision, then intervene in neurological diseases through means such as brain-computer interfaces, and finally explore the deep collaboration among the human brain, the digital twin brain and the cloud large model. 

"Today when we open our mobile phones to chat with large models such as DeepSeek, essentially there is a combination of human brain thinking and machine knowledge, but the interaction efficiency through typing or voice is actually very low." Professor Miao Feng made a vivid analogy. 

He mentioned that large language models and the human brain have their own strengths. If in the future, based on more efficient brain-computer interface, brain-inspired intelligence technology and even the individual's "digital twin brain", we can realize efficient direct connection and deep collaboration between the human brain, the twin brain and the cloud large model, and let "Human-in-the-loop" participate fully, we can form the human-machine integrated intelligence of "1+1>2". 

It can be seen that although bio-intelligence is still in the early stage of the technology curve, the three curves of demand, technology and capital have begun to intersect. 

How it collaborates and integrates with large models and embodied intelligence still needs product and market verification. What capital is positioning in advance is exactly this possibility of moving from marginal exploration to the mainstream intelligent architecture. 

Industry Practice: What Barriers Does Bio-intelligence Need to Cross to Move into the Industry?

It needs to be clarified that a blue ocean does not mean that technology will naturally grow into an industry. 

Bio-intelligence still needs to cross three barriers: forming joint forces through interdisciplinary integration, moving laboratory prototypes to stable products, and matching real scenario demands. 

The first problem to be solved is that interdisciplinary integration easily stays at "physical splicing" and it is difficult to form a continuous joint force for tackling key problems. 

Take Nanjing University as an example. Its layout in bio-intelligence has lasted for more than ten years. Six years ago, the team took the lead in breaking through the retina-like visual sensor and brain-inspired visual architecture in the world; in 2021, the Institute of Brain-inspired Intelligent Science and Technology was established. 

In recent years, Nanjing University has further strengthened the systematic layout of relevant cutting-edge fields. In 2026, under the overall planning and promotion of Academician Zheng Hairong, the Institute of Bio-Intelligence of Nanjing University was officially established. In this context, brain-inspired intelligence is an important part of bio-intelligence and can be regarded as a subset of it, and the previous brain-inspired research has laid a foundation for the Institute of Bio-Intelligence. 

Bio-intelligence spans the fields of physics, information, biology, medical engineering and others. To truly form scientific research joint forces, it requires scientists to break through the boundaries of existing disciplines, and also requires universities to provide support in discipline planning, talent introduction and cultivation, and scientific research platform construction. 

In this process, the concepts advocated by Academician Zheng Hairong, such as strengthening original innovation, encouraging cutting-edge exploration, and promoting interdisciplinary integration, have gathered interdisciplinary innovation forces and effectively promoted the development of scientific research. 

Specifically for the organizational structure of the institute, it adopts an interdisciplinary research design, relying on the idea of "shifting from science to engineering and landing in engineering" of the emerging engineering education, to make different disciplines collaborate around common technical goals, and promote bio-intelligence research from single-point technological breakthroughs to multi-disciplinary collaborative tackling of key problems. 

The Institute of Bio-Intelligence of Nanjing University focuses on four major research directions: 

1. Bio-intelligence fundamentals and large models: Analyze the cross-scale mechanism of life intelligence, and build interpretable, predictable and intervenable life models; 

2. Bio-inspired perception and computing: Develop bionic perception, sensing-computing integrated devices and neuromorphic computing; 

3. Biocomputing and organoid intelligence: Explore organoid living computing and wetware-hardware collaboration; 

4. Bio-intelligent agents and embodied systems: Build a closed loop of "perception - decision - action", and promote the integration of bio-intelligence and silicon-based intelligence. 

The four main lines advance layer by layer along "fundamental mechanism - key technology - system integration", forming a research chain from scientific discovery to intelligent entities. 

These research main lines have already spawned specific achievements. The team has applied the full-connection model of the fruit fly brain containing about 100,000 neuron connections to the UAV system, and realized the functions of escaping danger and avoiding obstacles in the laboratory without the need for traditional massive data training. In the future, it is expected to give birth to intelligent exoskeleton "mecha" that helps paralyzed patients walk, and also allow humans to precisely control entities across space with their minds just like in the movie *Avatar*. 

The second major difficulty is engineering: the performance advantages in papers must be turned into manufacturable, reusable products that can enter real systems. 

"Optical Speech Chip" is a sample for the team to try to cross this barrier. To understand its value, we first need to understand the difference between traditional machine vision and human vision. 

"For example, if I want to judge what brand the mobile phone in your hand is, if I use traditional computer vision processing methods, it needs to digitize 100% of all pixels of the entire picture and model all of them into images, including redundant background information, and then go through extremely energy-consuming data movement and back-end calculation to compare features. A lot of bandwidth and computing power are wasted on processing useless redundant information." Professor Miao Feng gave an example. 

The human eye relies on the "bipolar cells" of the retina to preprocess visual signals, and the cerebral cortex uses the attention mechanism to mobilize the "macular area" to focus on core features, completing rapid recognition with extremely low consumption. This prompted researchers to break the traditional path of "imaging first, processing later", and make the sensing, storage and computing deeply integrated at the front end. 

More than ten years ago, Professor Miao Feng's team started the cross-border exploration for bio-intelligence. Two months ago, the achievement of "Optical Speech Chip 1.0" was published in *Nature Sensors*, the top international journal in the sensing field, and *Nature Electronics*, the top international journal in the electronics field, also published a special article to recommend it. After the achievement was unveiled, the industry and capital circles have successively contacted the team. 

The upcoming "Optical Speech Chip 2.0" further shifts to architecture innovation oriented to engineering. The chip directly realizes "optical signal input, Token output" at the sensor end, so that optical signals can be directly connected to large language models without redundant conversion. At present, the chip has completed the engineering route, and is carrying out adaptation and industrial incubation for scenarios such as micro-special robots and commercial aerospace. Professor Miao Feng hopes that the Optical Speech Chip can become the entry point for large language models to perceive the physical world in the future. 

After crossing the engineering barrier, the third problem is how to realize the collaboration between talents and industries, so that technologies can continue to emerge and align with real demands. 

In response to this, the institute adopts the idea of "reversing the layout based on future industrial demands", sets up the interdisciplinary direction of "Bio-intelligent Science and Technology" and launches global recruitment, gathering top leading talents and engineering technical talents. 

In terms of industrial collaboration, Nanjing University has previously signed a comprehensive strategic cooperation agreement with CETC Hikvision Group, promoting enterprise scenario understanding and ecological capabilities to be embedded in scientific research in advance. Enterprises put forward problems, engineering teams