This article presents a novel approach to information processing using a single nonlinear dynamical node with delayed feedback, demonstrating its effectiveness in tasks such as speech recognition and time series prediction. The concept is based on 'reservoir computing', a machine learning paradigm inspired by the brain's ability to process information. Traditional reservoir computing requires a large number of interconnected nodes, but this study shows that a single nonlinear node with delayed feedback can achieve similar performance. The research demonstrates that delay-dynamical systems, even in their simplest form, can efficiently process information. This is achieved by using the nonlinear dynamics of a single node with delayed feedback to mimic the behavior of a traditional reservoir. The system is implemented using an electronic circuit, and the results show that it performs well in both speech recognition and time series prediction tasks. The study shows that the single nonlinear node can be used as a reservoir, with its dynamics influenced by its own output at a previous time. This system is easy to implement, requiring only two elements: a nonlinear node and a delay loop. The research also shows that the system can be used for dynamical system modelling, where it is trained to reproduce a certain signal. The results indicate that the single nonlinear node with delayed feedback can perform as well as traditional reservoirs, which typically require hundreds or thousands of nodes. This finding has implications for the development of resource-efficient technological implementations of reservoir computing. The study also highlights the potential of delay systems for information processing, suggesting that they could be used in various applications, including real-time processing and high-speed implementations. The research provides a simple and efficient method for information processing, which could lead to new developments in machine learning and artificial intelligence.This article presents a novel approach to information processing using a single nonlinear dynamical node with delayed feedback, demonstrating its effectiveness in tasks such as speech recognition and time series prediction. The concept is based on 'reservoir computing', a machine learning paradigm inspired by the brain's ability to process information. Traditional reservoir computing requires a large number of interconnected nodes, but this study shows that a single nonlinear node with delayed feedback can achieve similar performance. The research demonstrates that delay-dynamical systems, even in their simplest form, can efficiently process information. This is achieved by using the nonlinear dynamics of a single node with delayed feedback to mimic the behavior of a traditional reservoir. The system is implemented using an electronic circuit, and the results show that it performs well in both speech recognition and time series prediction tasks. The study shows that the single nonlinear node can be used as a reservoir, with its dynamics influenced by its own output at a previous time. This system is easy to implement, requiring only two elements: a nonlinear node and a delay loop. The research also shows that the system can be used for dynamical system modelling, where it is trained to reproduce a certain signal. The results indicate that the single nonlinear node with delayed feedback can perform as well as traditional reservoirs, which typically require hundreds or thousands of nodes. This finding has implications for the development of resource-efficient technological implementations of reservoir computing. The study also highlights the potential of delay systems for information processing, suggesting that they could be used in various applications, including real-time processing and high-speed implementations. The research provides a simple and efficient method for information processing, which could lead to new developments in machine learning and artificial intelligence.