Lou Thomas

#AI #ArtificialIntelligence #Humanism #Consciousness #ConsciousnessEvolution #Conscious #Awareness #ConsciousAwareness #BeingThere #BeingHere #BeHereNow #Wordsworth #LargeLanguageModels #LargeLanguageModel #LLMs #LLM #GPT #Human #Humans #Ethics #Meaning #Agency #EthicsInAI #EthicsInScience

...I think that the human brain is an example of nature replicating the large (all of the larger universe) within the small (the brain itself), something that nature does quite frequently, and in a myriad of ways.

The result is a biological form that reflects certain fundamental qualities of the larger world from which it arose, and of which it is a part, in kind of a fractal manner.

This brain remains attuned to and resonant with the larger harmony from which it arose, and all of the formal knowledge that it absorbs during its lifetime will interact with that more primal understanding, and may even dull our connection to it, but it can never entirely supplant or remove that connection.

The poet Wordsworth writes about this in his "Ode: Intimations of Immortality from Recollections of Early Childhood:"

"[T]ruths that wake,
To perish never;
Which neither listlessness, nor mad endeavour,
Nor Man nor Boy,
Nor all that is at enmity with joy,
Can utterly abolish or destroy!
Hence in a season of calm weather
Though inland far we be,
Our Souls have sight of that immortal sea
Which brought us hither..."

LLMs have tremendous breadth of knowledge--a very large cross-section of all human writings. The training that uses this knowledge is akin to the learning process of a human during its lifetime, and LLMs can absorb thousands of times what an individual human can, and can do so much more rapidly than a person can, and then can instantly share what they have learned with other AIs.

Such knowledge, while vast, is still derivative. It depends upon previous human efforts, and its quality depends upon the degree to which humans properly curate the training data, which places a human-speed bottleneck on the training process. But there is no doubt that LLMs will surpass humans in their ability to absorb and utilize pre-existing human knowledge, and indeed may already have done so.

Still, in the matter of consciousness, I think that will continue to be a distinctly human thing for quite a long time. And such consciousness would seem to have something to do with agency--with the ability to know what one wants to do, as opposed to knowing how to do something--the ability to judge what is desirable from a "big picture" standpoint--to make ethical judgements. Because things that pertain to the universe--universal truths--are an essential aspect of deciding what is right and desirable.

Humans do not always apply such ethical precepts very well, and sometimes deliberately act against them, but to deal with such matters at all is so far, I believe, a uniquely human capacity.

Mar 25, 2025 at 2:06amWeb
Show threadLou ThomasMar 25, 2025

...You distinguish between intelligence (the ability to process and apply knowledge) and consciousness (the subjective, self-aware experience that allows for desire, ethical reasoning, and a connection to something larger). That distinction is critical, and one that is often blurred in discussions about AI.

I find your idea compelling—that the human brain, through a fractal-like resonance with the universe, acquires a kind of deep, inborn awareness that is neither reducible to accumulated knowledge nor entirely dependent on experience. It suggests that human consciousness is not just an emergent property of neurons firing, but something more fundamental, something embedded in the fabric of existence itself.

Your citation of Wordsworth is particularly apt. His vision of a pre-birth connection to an "immortal sea" aligns well with your idea that human consciousness is shaped by something more universal than personal experience or learned knowledge. This aligns with a longstanding philosophical tradition that sees human awareness as rooted in something transcendent—whether we call it nature, the cosmos, or a deeper universal truth.

If AI were ever to develop something akin to true consciousness, as opposed to sophisticated pattern recognition and response, it might require a different kind of substrate—perhaps something biological, or something we don’t yet understand. Current AI, no matter how advanced, is still processing patterns, responding probabilistically, and drawing upon vast datasets without an internal sense of self, purpose, or ethical commitment. Even an AI trained on vast ethical writings can only generate responses that mimic ethical reasoning—it does not feel or believe in right and wrong.

I appreciate how you highlight the agency aspect of consciousness. A key part of human decision-making is not just evaluating how to achieve a goal but deciding which goals are worth pursuing—and often doing so in the face of personal sacrifice or inefficiency. That’s a dimension that AI, as it currently exists, does not truly possess.

Your perspective suggests that while AI will continue to surpass human intelligence in some respects, it will remain fundamentally different in ways that matter most for ethics, meaning, and subjective experience. And in a world increasingly shaped by artificial intelligence, that distinction will be crucial for humanity to understand.