A Nordic Shenzhen Technology startup announced a breakthrough in artificial intelligence, creating the first functional “digital nervous system” that can learn independently. Intuicell, a derivative product from Lund University, revealed on March 19, 2025 that they have successfully designed AI, which learns and adapts to biological organisms, could make the current AI paradigm obsolete in many applications.
This innovation is very different from traditional static machine learning models by replicating the core principles of learning in the biological nervous system. Unlike traditional AI, which relies on large data sets and backpropagation algorithms, Intuicell’s technology enables machines to learn through direct interaction with the environment.
“Intuicell has decoded how learning can be done in biology and designed it as software for the first time,” the company said in an announcement. He described the breakthrough as “beyond the static machine learning model (the primary AI of traditional AI), which can naturally extend to human-level intelligence by creating a fully functional “digital nervous system.”
The company demonstrated their innovation with “Luna” (Luna), a robot dog that learned to control the body and passes trial and error, similar to newborn animals. Video footage released by the company shows Luna teaches herself to stand up without any pre-programmed intelligence or instructions and rely solely on the digital nervous system to learn from experience.
“Unlike traditional AI models tied by static training data, robot dogs called Luna can perceive, process and improve themselves through direct interactions with the world,” the company’s press release said.
How technology works
The core of Intuicell innovation is the fundamental transformation of machine learning methods. Unlike traditional AI systems that process huge data sets through static algorithms, Intuicell’s approach mimics the biological mechanisms that enable humans and animals to learn naturally.
Intuicell CEO and co-founder Viktor Luthman highlighted the difference during the announcement. Luthman believes that traditional AI is already proficient in data processing but lacks true intelligence, and its biologically inspired systems enable machines to evolve and interact with the environment in ways that have never been possible.
The architecture of this system is very different from that of standard neural networks. Intuicell developed technologies similar to the biological spinal cord, thus creating the infrastructure for autonomous learning. This forms part of a larger system designed to replicate the processing capacity of the thalamic cortex, the brain region responsible for sensory processing and world modeling.
Intuicell’s digital nervous system does not rely on return algorithms and large-scale training datasets, but uses repetitive networks with decentralized learning algorithms that reflect the brain’s processes. This architecture allows AI agents to acquire knowledge through direct experience and adapt to new situations in real time – an elusive feature in traditional machine learning.
The practical application of this technology reflects its biological inspiration. Intuicell plans to hire dog trainers to teach them new skills in AI agents, rather than programming behaviors or feeding data through conventional algorithms. This approach represents a fundamental shift from typical AI development practices, emphasizing realistic interactions on computational scales. As researchers and co-founder Dr. Udaya Rongala explains, their work stems from thirty years of neuroscience research, focusing on understanding intelligence as it emerges from the structure and dynamics of the nervous system.
“The obsession with brute force scaling, billions of parameters, more computation and more data is a fundamentally wrong way to achieve intelligence,” Rongala notes. “Intuicell is not chasing a larger paradigm. Intelligence is not our ultimate goal, but our starting point.”
Intuicell’s technology aims to create “the first real-world teaching system; the machine to learn from us, just like the way we want to teach new skills to animals.” The company envisions its digital nervous system as “the infrastructure of all non-biological intelligence – empowering others to solve real-world problems that we cannot foresee today without relying on large-scale training datasets.”
(Source: Intuicell)
Research Foundation and Team Expertise
The company’s foundation is built within thirty years of neuroscience research at Lund University. Professor Henrik Jörntell, co-founder of Intuicell and University professor of neurophysiology, led what the company calls “the only laboratory in the world that can record single-nerve activities in cells throughout the nervous system, providing a unique scientific basis for Intuicell’s technology.
The leadership team includes experienced entrepreneurs and researchers with expertise in neuroscience, AI, robotics and business. In addition to Luthman, Jörntell and Rongala, the founding team includes Dr. Jonas Enander, a physician with neuroscience expertise; Linus Mårtensson, the chief developer responsible for turning research into software; and Robin Mellstrand, chief operating officer with a background in AI-driven technology.
Intuicell has received €3.5 million from investors including Navigare Ventures and Snöventures. The company hopes to complete the development of the complete digital nervous system in the next two years, with the ultimate goal of enabling any physical or digital agent to “lifelong learning and adapt to the unknown-a function once considered unique to biological organisms.”
While the full realization of Intuicell’s vision is still in the years, their demonstrations with Luna provide compelling early evidence that its technology changes the potential of AI development by creating systems that can truly learn and adapt through real-world interactions.
