Recent advancements in artificial intelligence, particularly in chatbot technology, have led to responses that can feel remarkably personalized, creating an impression of interaction with a conscious entity. However, researchers from the University of Montreal and Johns Hopkins University emphasize the need to distinguish between intelligence and consciousness.
The study references the phenomenon known as “blindsight,” observed in individuals with damage to the primary visual cortex. These patients report an inability to see objects in their field of vision yet can accurately guess the direction of movement or the emotional expressions of faces in photographs.
Vanessa Hadid, a psychology researcher from Montreal, explains, “A person with blindsight successfully processes and utilizes visual information but lacks conscious experience of seeing itself.” This example illustrates that complex information processing and accurate responses do not necessarily indicate the presence of an inner world or awareness.
Current language models function as purely computational systems. They produce coherent and contextually appropriate responses through statistical analysis of vast datasets, rather than through personal feelings or life experiences.
As algorithms increasingly mimic human behavior, users may fall into the trap of anthropomorphism—a psychological tendency to attribute emotions and intentions to inanimate objects. This creates a misleading impression that machines “understand” humans, fostering unwarranted trust in their capabilities.
This risk is particularly pronounced when individuals are in states of emotional vulnerability, stress, or loneliness. They may seek psychological support from chatbots, confusing mere textual comfort with genuine care.
Hadid warns, “The primary danger lies not in the possibility of AI providing incorrect advice, but in its potential to respond so effectively that we forget the crucial point: there is no one behind that response.”
Researchers clarify that their work is not intended to discourage the use of AI technologies. Instead, it serves as a call for more informed and pragmatic engagement with these tools.
Understanding that neural networks lack moral frameworks, empathy, or inner experiences is essential for proper interaction—viewing them as sophisticated and powerful tools. While algorithms can organize information or assist with routine tasks, they can never replace genuine human relationships, authentic mutual support, or professional mental health care from living practitioners.
The study highlights the distinction between intelligence and consciousness in AI, warning against the anthropomorphism of chatbots. It emphasizes the importance of recognizing the limitations of AI in providing emotional support.