A recent study conducted by scientists at Skoltech raises intriguing questions about the number of senses humans possess and how this understanding could influence advancements in robotics. The research introduces a mathematical model that explores the encoding and storage of memory, suggesting an optimal configuration of seven dimensions in conceptual space, akin to having seven senses.
The findings imply that both human cognition and artificial intelligence systems could enhance their learning and recall capabilities by integrating a broader range of sensory inputs. This could significantly benefit the development of robotic systems, improving their functionality and interaction with the environment.
According to the study, when concepts in memory are characterized by seven distinct features, as opposed to fewer or more, the capacity for holding distinct objects in memory is maximized. This research centers on “engrams,” which are the physical representations of memory formed by groups of neurons that fire together. Each engram corresponds to a concept, articulated through a set of sensory features.
The concept of a banana, for instance, encompasses its visual appearance, scent, flavor, and other sensory attributes, effectively making it a five-dimensional entity within the mental landscape of the brain. These engrams evolve over time, becoming either more defined or more diffuse based on the frequency of sensory stimulation they receive from the external world.
Lead researcher Nikolay Brilliantov commented on the findings, stating, “We have mathematically demonstrated that the engrams in the conceptual space tend to evolve toward a steady state, meaning that after an initial period, a stable distribution of engrams emerges that remains constant over time.” This observation underscores the notion that seven dimensions in conceptual space yield the greatest number of distinct engrams stored in memory.
Brilliantov elucidates that this dimension does not depend on the specific characteristics of the conceptual space or the nature of sensory stimuli. Instead, the number seven appears to be a consistent and enduring characteristic of memory engrams. One important consideration is that multiple engrams that are similar and cluster around a common center are counted as one in assessing memory capacity.
The complexity of human memory, intertwined with consciousness, remains a subject of great intrigue. Enhancing theoretical models of memory can lead to deeper insights into human cognition and aid in replicating human-like memory in artificial intelligence systems. The study”s results are detailed in the journal Scientific Reports, under the title “The critical dimension of memory engrams and an optimal number of senses.”
