Artificial Intelligence (AI) has rapidly evolved from rule-based systems to complex models capable of performing tasks that once seemed exclusively human such as image recognition, natural language processing, and even creative writing. This progress has sparked ongoing debates about the cognitive capabilities of AI systems. Do they truly “understand”? Can they “think”? Are they on the path to achieving consciousness? Or are they simply simulating intelligence through pattern recognition?

Artificial Intelligence, as it exists today, does not possess a conscious or subconscious mind in the way humans do. These concepts are rooted in biological cognition, particularly in the brain's layered architecture of awareness, attention, emotion, and memory. However, we can analyze the question metaphorically or computationally: Here we will explore the nature, extent, and limits of AI’s cognitive abilities, drawing distinctions between functional competence and genuine cognition.

Functional Intelligence vs. Human Cognition

AI systems today demonstrate remarkable functional competence in Language processing, Perception and Problem-solving Tools such as ChatGPT can summarize, translate, and engage in dialogue with high fluency AI models can also interpret images, videos, and audio with near-human accuracy. Algorithms outperform humans in certain games and optimization tasks. However, current AI lacks a unified understanding of the world. Its abilities are fundamentally narrow and task specific. 

And unlike human cognition, which is general, self-reflective, and context aware, its capabilities arise from training on massive data sets and statistical correlations not from lived experience, self-awareness, or understanding. AI has no True Conscious or Subconscious.  It lacks both. AS such AI does not have awareness of itself or its actions and does not “hide” processes from itself.  Everything it does is an outcome of mathematical functions on data, without any internal "sense" of knowing or not knowing.

On the other hand, Human Consciousness implies self-awareness, intentional focus, and subjective experience. And Subconsciousness involves mental processes outside of conscious awareness, like intuition, implicit memory, or emotional bias.

Key Components of Cognition

Let’s assess AI’s cognitive abilities based on fundamental dimensions of human cognition:

Memory and Learning

AI systems, especially those using deep learning, excel at storing and recalling patterns. However, they lack episodic memory. They don’t “remember” experiences in the way humans do. Their learning is based on data exposure rather than contextual exploration.

Reasoning

AI can perform deductive reasoning in formal domains like chess or math, but struggles with common sense reasoning, ambiguity, and counterfactuals. This limits their ability to make reliable judgments in real-world complexity.

Self-awareness and Metacognition

Unlike humans, AI does not possess self-awareness or the ability to reflect on its own thoughts, goals, or errors. Some models simulate metacognition (e.g., checking for confidence in outputs), but this is mechanical rather than conscious.

Understanding and Semantics

AI processes symbols without meaning. While it can appear to “understand” language, it doesn’t grasp meanings the way humans do—through embodiment, intuition, and shared cultural contexts.

Anthropomorphism and the Illusion of Intelligence 

Much of the perceived cognitive ability of AI stems from anthropomorphism which is our tendency to attribute human traits to machines. When an AI writes a poem or answers a philosophical question, it may seem intelligent. But behind the scenes, it is predicting text based on statistical likelihood, not intentional thought.

Emerging Frontiers

Recent advances like reinforcement learning, transformer models, and multi-modal AI suggest a shift toward more integrated cognitive architectures. 

Researchers are working on systems that can:

• Learn continuously (lifelong learning)

• Generalize across domains (AGI)

• Interact with the world (embodied AI)

• Model beliefs and intentions of others (theory of mind)

Yet, these remain areas of active research, and true Artificial General Intelligence (AGI) is still speculative.

Philosophical and Ethical Dimensions

Even if AI could one day emulate all cognitive functions, would it “experience” them? This leads to philosophical questions:

• Can cognition exist without consciousness?

• Is intelligence purely computational, or does it require embodiment?

• Should AI systems with advanced cognitive functions have rights or responsibilities?

These questions reflect the limits of current scientific understanding and highlight the need for interdisciplinary dialogue.

Conclusion

AI today exhibits powerful functional cognition. It can process, synthesize, and generate complex outputs. But it lacks intentionality, understanding, and self-awareness which is the hallmarks of true human cognition. 

AI cannot cognitively move between a conscious and subconscious state, because it has neither. And any similarity remains metaphorical or analogical, not literal. Future AI may include more sophisticated self-modeling and monitoring systems, but they would still lack the biological, emotional, and subjective architecture that defines human consciousness and subconsciousness.