Louis Longin

People won’t hold cars responsible for traffic accidents, yet they do when artificial intelligence (AI) is involved. AI systems are held responsible when they act or merely advise a human agent. Does this mean that as soon as AI is involved responsibility follows? To find out, we examined whether purely instrumental AI systems stay clear of responsibility. We compared AI-powered with non-AI-powered car warning systems and measured their responsibility rating alongside their human users. Our findings show that responsibility is shared when the warning system is powered by AI but not by a purely mechanical system, even though people consider both systems as mere tools. Surprisingly, whether the warning prevents the accident introduces an outcome-bias: the AI takes higher credit than blame depending on what the human manages or fails to do.

Perception often occurs in the presence of others. Here we introduce joint perception as what occurs when the perceptual states of perceivers get mutually entangled, irrespective of the occurrence of joint attention. From a conceptual perspective, this entanglement results in two or more perceivers being aware that objects or scenes are mutually seen and at least visible. From an empirical perspective, this entanglement also results in a difference in how the perceivers see the objects or scenes in question. Here we argue that the empirical evidence of a social sensitivity of perception is best explained by appeal to visual representations. We propose that joint perception is a place where visual representations play both an explanatory and constitutive role. We still point that representationalism needs to better explain what happens to the subject of perception when ‘we see’ an object, instead of seeing it alone.

Humans and other animals possess the remarkable ability to effectively navigate a shared perceptual environment by discerning which objects and spaces are perceived by others and which remain private to themselves. Traditionally, this capacity has been encapsulated under the umbrella of joint attention or joint action. In this comprehensive review, we advocate for a broader and more mechanistic understanding of this phenomenon, termed co-perception. Co-perception encompasses the sensitivity to the perceptual engagement of others and the capability to differentiate between objects perceived privately and those perceived commonly with others. It represents a distinct concept from mere simultaneous individual perception. Moreover, discerning between private and common objects doesn't necessitate intricate mind-reading abilities or mutual coordination. The act of perceiving objects as either private or common provides a comprehensive account for social scenarios where individuals simply share the same context or may even engage in competition. This conceptual framework encourages a re-examination of classical paradigms that demonstrate social influences on perception. Furthermore, it suggests that the impacts of shared experiences extend beyond affective responses, also influencing perceptual processes. This article is categorized under: Psychology > Attention Philosophy > Foundations of Cognitive Science Philosophy > Psychological Capacities.

Artificial Intelligence (AI) profoundly affects how people communicate, work, and perceive the world. While autonomous AI systems are the focal point in societal and academic discussions, advisory AI systems, which influence human decisions but don't undertake independent actions, often remain unexplored. Examples range from automated purchase recommendations to medical diagnoses. This dissertation seeks to understand what advisory AI systems truly are. Are they capable of autonomous, human-like action? Or can they be reduced to inert tools? And what happens when advisory AI systems are closely linked with human perception, especially through Augmented Reality and sensory augmentation? Does their ontological status change? This dissertation concludes that, regardless of their implementation, advisory AI systems occupy an ontological status between tools and humans. They are more than just tools but less than humans.

The recent rise of artificial intelligence (AI) systems has led to intense discussions on their ability to achieve higher-level mental states or the ethics of their implementation. One question, which so far has been neglected in the literature, is the question of whether AI systems are capable of action. While the philosophical tradition appeals to intentional mental states, others have argued for a widely inclusive theory of agency. In this paper, I will argue for a gradual concept of agency because both traditional concepts of agency fail to differentiate the agential capacities of AI systems.