The following is a mini-dose of a philosophical style stream of consciousness about what it means to have meaning in action. It is dense and short and definitely deserves elaboration at a later time.
Motivation and goal-directed behavior arise from goal-setting, conditional environmental stimuli, and internalized mental models shaped by experience and culture (bias). To be biased may be politically incorrect, but bias in itself is core to conscious living. These cognitive processes are dynamically structured and modulated through working memory as interaction, time, and environmental engagement goes on. These adhere to and shape bias, tempted by the self, which directs and distributes attention. All of this contributes to sentience and a way to afford task completion and the formation of meaning through manipulating and being manipulated by our environment.
In contrast, contemporary Large Language Models (LLMs) operate through statistical pattern recognition, primarily via next-token prediction over large collections of human-generated text. While they can generate syntactically coherent and contextually relevant outputs, they lack intrinsic mechanisms for working memory, long-term intentional planning, or genuine semantic comprehension. Apparent coherence in conversation is often the product of shallow contextual continuity within a fixed token window, rather than the product of a stable, working memory like that of humans.
Although artificial neural networks are loosely inspired by biological neurons and can exhibit forms of pattern completion through backpropagation, but they remain flowed. In that they do not possess the embodied, culturally shaped, interactive and evolving bias through which humans develop meaning.
Human cognition is shaped by lived interaction with the world; in society, culture, and history, and the ability to experiences and engage with perception, emotions, and symbols. In contrast, LLMs lack such grounding and operate without the intentionality or the subjectivity that characterizes human thought.
We understand consciousness as the lack thereof. And there is no guarantee that a "more" conscious being would consider us to be conscious because, while we exhibit a level of intelligence, consciousness encompasses and requires more than intelligence. So could LLMs be exhibiting or AI, one day, exhibit a lesser form of consciousness that we don't recognize?
Perhaps.
While LLMs can simulate aspects of linguistic competence, they do so without the intention, depth, or the commitments that underlie genuine understanding.
This is the distinction between apparent fluency and actual cognition.
Key Takeaways:
- Human motivation emerges from bias and intention: Goal-directed behavior arises from environmental stimuli, cultural experience, and internalized mental models. Bias is not just a flaw, but a necessary feature of sentient cognition.
- Bias directs attention and action: It shapes our perception, fuels our drive, and evolves through interaction with time, context, and environment modulated by working memory and self-awareness.
- LLMs simulate coherence without comprehension: They rely on statistical next-token prediction within limited context windows, not intrinsic working memory or intentional planning.
- Human cognition is embodied, situated, and social: Meaning arises from lived experience, history, emotion, and symbolic interpretation these are things machines do not, currently, possess.
- Artificial Intelligence lacks cultural grounding: Though neural networks are inspired by biology, they do not experience or evolve through context, bias, or shared human meaningfulness.
- Fluency is not cognition: LLMs can appear fluent, but lack depth, intent, and commitment—fundamentals of actual understanding.
- Consciousness exceeds intelligence: A system may appear intelligent without being conscious. Likewise, consciousness in others, whether less or more than ours, may be unrecognizable to us.
- Apparent understanding is not real understanding: The difference between human and artificial cognition lies in the presence (or absence) of meaning, not just linguistic form.