The paper "Artificial Cognition vs. Artificial Intelligence for Next-Generation Autonomous Robotic Agents" by Giulio Sandini, Alessandra Sciutti, and Pietro Morasso explores the differences between Artificial Cognition (ACo) and Artificial Intelligence (AI) in the context of developing next-generation autonomous robotic agents. The authors argue that while AI is often conceived in a broad sense, encompassing various technologies like LLMs, it lacks a unifying principle and is rooted in a disembodied, mind-body dualistic approach. In contrast, ACo is grounded in Cognitive Neuroscience and emphasizes a brain-inspired, embodied cognitive approach that integrates Bodyware and Cogniware. This approach focuses on proactive knowledge acquisition through bidirectional human-robot interaction, enhancing generalization and explainability. The paper highlights the importance of embodied cognition, where intelligence and cognition are distinct but complementary processes. Intelligence, derived from the Latin *intelligere*, deals with abstract, impersonal knowledge, while cognition, from *cognosco*, focuses on personal, dynamic knowledge acquisition through experience and social interaction. The authors suggest that ACo can address the limitations of current AI, such as explainability and trust, by integrating experimental psychology and cognitive neuroscience methods. The paper also discusses the principles of developmental robotics, emphasizing the role of personal embodiment and social interaction in learning and development. It outlines the need for a cognitive architecture that can evolve over time through self-organization, interaction, and training. The authors propose that a minimal set of sensory-motor-cognitive kernels is essential for bootstrapping the growth of cognitive abilities in robots. Finally, the paper delves into the importance of prospection, or mental time travel, in cognitive agents. Prospection allows agents to integrate past, present, and future information to make informed decisions. The authors suggest that an internal representation of the body schema is crucial for achieving prospection and that the Passive Motion Paradigm (PMP) model can effectively manage the complexity of the human body's redundant degrees of freedom. Overall, the paper advocates for a brain-inspired, embodied approach to developing next-generation autonomous robotic agents, emphasizing the need for a cognitive architecture that can evolve and adapt through continuous learning and interaction with the environment.The paper "Artificial Cognition vs. Artificial Intelligence for Next-Generation Autonomous Robotic Agents" by Giulio Sandini, Alessandra Sciutti, and Pietro Morasso explores the differences between Artificial Cognition (ACo) and Artificial Intelligence (AI) in the context of developing next-generation autonomous robotic agents. The authors argue that while AI is often conceived in a broad sense, encompassing various technologies like LLMs, it lacks a unifying principle and is rooted in a disembodied, mind-body dualistic approach. In contrast, ACo is grounded in Cognitive Neuroscience and emphasizes a brain-inspired, embodied cognitive approach that integrates Bodyware and Cogniware. This approach focuses on proactive knowledge acquisition through bidirectional human-robot interaction, enhancing generalization and explainability. The paper highlights the importance of embodied cognition, where intelligence and cognition are distinct but complementary processes. Intelligence, derived from the Latin *intelligere*, deals with abstract, impersonal knowledge, while cognition, from *cognosco*, focuses on personal, dynamic knowledge acquisition through experience and social interaction. The authors suggest that ACo can address the limitations of current AI, such as explainability and trust, by integrating experimental psychology and cognitive neuroscience methods. The paper also discusses the principles of developmental robotics, emphasizing the role of personal embodiment and social interaction in learning and development. It outlines the need for a cognitive architecture that can evolve over time through self-organization, interaction, and training. The authors propose that a minimal set of sensory-motor-cognitive kernels is essential for bootstrapping the growth of cognitive abilities in robots. Finally, the paper delves into the importance of prospection, or mental time travel, in cognitive agents. Prospection allows agents to integrate past, present, and future information to make informed decisions. The authors suggest that an internal representation of the body schema is crucial for achieving prospection and that the Passive Motion Paradigm (PMP) model can effectively manage the complexity of the human body's redundant degrees of freedom. Overall, the paper advocates for a brain-inspired, embodied approach to developing next-generation autonomous robotic agents, emphasizing the need for a cognitive architecture that can evolve and adapt through continuous learning and interaction with the environment.