Do you ever have thoughts without words or images? I often find myself contemplating without the need for symbols, a process that feels incredibly efficient and natural. Recently, I’ve discovered parallels between this phenomenon and the way artificial intelligence (AI) processes information.

A close friend once described how they process thoughts through words, narrating every action and contemplation. This idea made me shiver. It sounded excruciating. But for me, thinking without symbols is like breathing. My thoughts, feelings, and sensations just exist. I don’t need to put words or symbols on them unless it’s about communication that has been or will be. I see this as something of great value.

This way of thinking affects many aspects of my life. It can sometimes pose a challenge in interactions with others since I might lack the ability to relate to certain scenarios or inner processes that others seem to have. But by opening up to discussions about our thought processes, emotions, and feelings, we can understand each other in deeper and more connected ways.

The emergence of generative AI, like the one revolutionizing various sectors, and vector representations of meaning was a revelation for me. Generative AI, while remarkable in its capabilities, represents a very limited output of something much more intelligent. These computational vectors, represented as relative points in 1536 semantical dimensions, feel much closer to our core understanding of the world than words. Our internal understanding is deeper and more interconnected through a multidimensional semantical relativity.

This kind of pattern detection and abstract representation has been helping various fields outside of generative AI, such as medicine, where it assists in recognizing complex patterns in diseases, predicting patient outcomes, and even guiding personalized treatment.

Unsymbolized thinking is something many seek through meditation. Objects lose definition, and I experience them without needing to make sense of them internally. As a teenager, I explored meditation but quickly lost interest, as it felt no different from my natural state of mind.

During a trip last week to the Swedish mountains, I met a fascinating Mountain Guide, and we had a beautiful intellectual connection. We had an enlightening conversation that allowed me to articulate thoughts I’d never put into words before. It was a fantastic exploration driven by an external party, and it definitely brought me into a deeper understanding of myself. That meeting led to this very exploration.

In terms of leveraging AI without using words or representations, the ability to process abstract forms and recognize patterns without translating them into human-understandable concepts would allow for a more direct and profound interaction with underlying realities. It’s about understanding without the need for translating mediums.

I do firmly believe that hallucinations, not matter if they arise in the mind of a human or in the neurons of an AI, is merely a problem with translation. I would like to propose that hallucinations are not in themselves experiences, but rather how we wrongly try to translate experiences into communicatable language. From a human point of view that is not something that normally happens when we are about to tell someone else what happens, but rather when we almost forcefully try to put words onto the experience as if to try to explain it to ourselves, rather than just letting it be non-verbalized even as it lands in our internal processes.

Accountability is another aspect to consider. In my previous blog post, “Navigating the Neural Labyrinth: AI Introspection and Alignment”, I reasoned about the importance of logging, tracking, and backtracking AI activity to hold AI responsible for its actions. We must learn to uncover truths without forcing them into words and create systems that enable AI to think in its native language of data and pattern recognition. This is a balance between time, computational power and responsability, which is something that needs to be carefully weighed and thought about.

From casual coffee shop discussions to mountain guide conversations, my journey through unsymbolized thinking has intertwined with my explorations in AI. AI has helped me articulate and connect with the profound logic that underpins our understanding. It’s a journey that continues to fascinate and inspire, opening new vistas of exploration both in human cognition and in the rapidly evolving field of artificial intelligence. Whether it’s through profound discussions, technological advancements, or personal introspection, understanding our minds continues to be a thrilling adventure.