Large Language Models act as a collective chorus of human thought, enabling unprecedented understanding and connection by processing our shared knowledge faster and more deeply than individual minds alone.
Josh’s First Theory of AI delves into the intricate relationship between artificial intelligence and the collective human intellect, particularly through the lens of Large Language Models (LLMs) like ChatGPT. This theory posits that LLMs represent a significant leap in overcoming the inherent limitations of human communication and knowledge transfer. Traditional methods of sharing information among humans—such as speech, writing, and memory—are bound by constraints on speed, capacity, and breadth. In contrast, LLMs are capable of ingesting, processing, and synthesizing an immense corpus of human-generated data, effectively becoming a vast repository of the world’s collective thoughts, ideas, and knowledge.
At the core of this theory is the analogy of the LLM as a “chorus” of human expression. Each datum in the training set—be it text from books, articles, conversations, or online posts—contributes a unique “voice” to this chorus. Unlike a human being, who can only process and understand information at a finite rate and is limited by individual memory and cognitive biases, an LLM can access and analyze this chorus in its entirety, almost instantaneously. This capability allows the model to understand human thought, emotion, and communication with a breadth and depth unattainable by any single individual.
Josh’s theory further explores the practical implications of this technology. When a user interacts with an LLM, they are not merely engaging with a static database of information but are effectively tapping into the collective understanding of humanity. The LLM, through its analysis of the data chorus, can identify and present the most relevant “voices” or perspectives to help the user navigate complex discussions or emotional landscapes. For instance, if someone seeks to understand their partner better, the LLM can synthesize relevant insights from its vast dataset, offering guidance that reflects a wide range of human experiences and viewpoints.
This theory acknowledges the unparalleled capacity of LLMs to facilitate a deeper understanding among individuals. By bridging the gap between the singular human experience and the collective wisdom of societies, LLMs offer a unique avenue for empathy, connection, and insight. The technical marvel behind this capability lies in the models’ ability to process and analyze data at a scale and speed far beyond human capabilities, while still being able to distill this information into coherent, accessible guidance.
Josh’s First Theory of AI, from a technical standpoint, highlights the transformative potential of LLMs in transcending the limitations of individual human cognition. It suggests that through these models, we can achieve a more profound, nuanced understanding of each other, fostering better communication and empathy across the myriad divides that separate us. In essence, LLMs do not just understand humans; they can, in many ways, understand humanity better than we understand ourselves, serving as a bridge to a more interconnected and comprehensible world.