Magnus Johnsson

AI, Cognition, and Data-Driven Innovation

Research: Exploring AI, Cognition, and Beyond

Understanding the brain and cognition Thinker

I am a cognitive and computer scientist conducting interdisciplinary research at the intersection of artificial intelligence (AI), cognitive science, and philosophy. My work focuses on hybrid intelligence, integrating human intuition with AI to address complex challenges in domains like financial markets, medical diagnostics, and sustainable energy, together with more theoretical and philosophical work. Detailed on my Publications page, my research advances our understanding of biological and artificial intelligence while delivering innovative solutions to real-world problems.

Key Research Areas

Hybrid Intelligence

Illustration of hybrid intelligence

I explore hybrid intelligence by combining human insights, such as geopolitical or domain-specific expertise, with advanced AI algorithms like self-organizing maps (SOMs) and associative self-organizing maps (ASOMs). This approach includes aspects such as explainable AI and fluent interaction between natural and artificial cognitive systems to obtain optimal performance of the hybrid system with humans in the loop. This enhances decision-making in dynamic environments, from financial trend prediction—blending quantitative analysis with human strenths such as qualitative judgment and intuition—to risk management and strategic speculation, offering both academic insights and practical tools.

Medical and Health Applications

Illustration of medical AI DNA

In healthcare, I leverage AI to improve diagnostics and treatment, notably in in-vitro fertilization (IVF) by optimizing embryo selection through machine learning. Collaborating with researchers like David Gil at the University of Alicante, and researchers from Malmö University, and Kristianstad University, I develop decision-support systems for fields like urology, Parkinson’s, and fertility, combining neural networks with clinical expertise to enhance outcomes.

Financial Market Applications

Illustration of financial applications

I apply AI and hybrid intelligence to financial speculation, using SOMs and recurrent neural networks (RNNs) to analyze time-series data from equities and commodities. By incorporating geopolitical and macroeconomic factors, my models enhance predictive accuracy, offering practical strategies for speculation and risk management while contributing to academic research.

Cognition and Consciousness (Biological and Machine)

My research on consciousness, in the phenomenological sense of subjective experience, spans biological and artificial systems. Using integrated information theory (IIT), I model consciousness through neural networks like SOMs and convolutional neural networks (CNNs) to study emergent cognitive processes. My work on cognition informs market psychology, cognitive biases in financial decision-making, and the development of ethically grounded AI with human-like awareness.

Action Recognition

I develop AI systems for action recognition, using hierarchical SOMs, CNNs and RNNs to interpret human actions from visual data. Collaborating with researchers from the University of Palermo and Lund University, I’ve created hierarchical SOM-based architectures to recognize actions and predict intentions, with applications in human-robot interaction, surveillance, and healthcare.

Imagination in Artificial Systems

Illustration of artificial imagination

I investigate imagination in AI, enabling systems to generate and manipulate internal representations of hypothetical scenarios. Inspired by the neuroscientific internal simulation hypothesis, my neural network models simulate mental imagery and predictive planning, advancing robotics, autonomous systems, and simulation-based training.

Associative Self-Organizing Maps (ASOM)

I invented the Associative Self-Organizing Map (ASOM), a neural network that integrates self-organization with associative learning to process multimodal data. This versatile framework enhances object recognition and sensory integration in robotics, advancing computational intelligence and cognitive modeling.

Neural Networks and Computational Intelligence

My work with neural networks, including SOMs, CNNs, RNNs, and my novel Tensor Multiple Peak SOM (T-MPSOM), focuses on emulating biological cognitive processes. These models drive innovation in pattern recognition, sensory processing, and cognitive modeling across financial, medical, and robotic applications.

Cognitive Robotics

I have designed cognitive robotic systems, such as the LUCS Haptic Hand series and LUCS Arm 1, to study haptic perception and shape recognition. Though my focus has shifted to cognitive modeling, these systems provided foundational insights into embodied AI and sensory integration.

Haptic Perception

My research in haptic perception explores active tactile exploration in robots using sensors and neural networks like the T-MPSOM. This work, which classifies objects by shape and texture, has applications in robotics, prosthetics, and human-machine interfaces.

Materials Science and Fusion Reactors

Together with researchers at Malmö University, I'm involved in a project that applies machine learning to simulate material degradation in fusion reactors, focusing on neutron-induced embrittlement to improve reactor design and sustainability. This interdisciplinary work showcases AI’s potential in addressing complex materials science challenges.

Other Research Contributions

My portfolio includes multimodal sensory integration, cognitive modeling, and contributions to the IKAROS project, a framework for cognitive robotics and brain simulation. My simulations of somatosensory reorganization, in collaboration with medical professionals, have informed strategies for optimizing recovery after nerve injuries.

Current and Future Directions

Illustration of current and future directions

My ongoing projects reflect my commitment to advance AI and cognitive science. In addition to my work on IVF, fusion reactor materials, financial market applications and other applications of hybrid intelligence, I continue to explore the theoretical foundations of consciousness and imagination in artificial and natural systems. My research aim to develop robust, generalizable AI models capable of operating in complex, dynamic environments. By integrating insights from philosophy, neuroscience, and computer science, I aim to develop robust, generalizable AI systems for complex, dynamic environments, while keeping humans in the loop.

Interested in my research?