Consider the realm of relational artificial intelligence, such as ChatGPT-4, and its implications. The successes and failures of advanced AI systems have sparked discussions on the extent and dangers of such technology. However, what insights can be gleaned from this research into the cognitive qualities akin to human cognition?
Throughout my career as a scientist and cognitive researcher, I have delved into understanding the intricacies of the human brain. Drawing from fields like psychology, neuroscience, and artificial intelligence, I have uncovered parallels and disparities between the operational mechanisms of healthy minds and conceptual AI systems. By exploring these distinctions, we can gain a deeper understanding of both entities.
These systems, like ChatGPT, are equipped with a relational framework that empowers them to predict patterns in data or communication. By discerning underlying patterns within datasets, they can generate plausible variations of such information autonomously, a characteristic that defines their relational nature. For example, ChatGPT can craft compelling narratives based on a vast repository of textual data, responding to user queries with engaging content. While ChatGPT excels in processing words, there exist AI models specialized in other domains like visual art, capable of generating original pieces reminiscent of renowned artists such as Picasso.
How does this relate to the human brain? The human mind has evolved to interpret diverse sensory inputs, primarily visual stimuli, to navigate and thrive in our environment. Crucially, our brains must anticipate the consequences of our actions on our sensory perceptions within the context of survival and adaptation. This dynamic interplay between actions and subsequent sensory feedback shapes our cognitive processes and decision-making.
The educational value of this cognitive framework lies in distinguishing between mere associations and causal relationships. Understanding the causal link between actions and outcomes is pivotal for effective decision-making and goal achievement. Unlike AI systems that focus on predicting specific data sequences, human brains are finely attuned to the consequences of actions in shaping our perception of reality.
In contrast to the intricate interplay of actions and perceptions in human cognition, AI models like ChatGPT primarily serve as predictive models for textual outputs rather than holistic representations of the world. While they provide insights into textual trends and patterns, they lack the experiential dimension of actively engaging with and influencing the physical world. This functional limitation underscores the fundamental disparity between AI-generated predictions and human cognition grounded in real-world interactions and consequences.
The potential for AI systems to evolve into proactive agents capable of interacting with their environment and observing the outcomes of their actions raises intriguing possibilities. By closing the loop between action and observation, AI models could transcend their current predictive capabilities and engage in empirical learning akin to human experiential knowledge acquisition.
However, essential aspects such as internal physiological states and the profound impact of social practices on individual experiences pose significant challenges for AI systems in replicating human-like cognition. The intricate interplay between physiological signals, social interactions, and cognitive processes underscores the complexity of human cognition that transcends mere predictive capabilities.
In essence, the distinctiveness of human cognition lies in our capacity for self-reflection, knowledge acquisition, and belief formation, attributes that have propelled us to innovate and evolve as a species. While AI systems may emulate certain cognitive functions, the essence of human knowing and understanding remains unparalleled in its depth and complexity.
As we contemplate the future trajectory of AI development, the prospect of AI systems evolving into self-sustaining entities with the capacity for cultural, social, and existential growth raises profound questions about the nature of artificial intelligence and its potential to mirror or diverge from human cognitive paradigms.
