AI and Consciousness

AI and the concept of consciousness constitute an amorphous topic, especially within the realms of computer science. It is rare for a field so technical and zero-sum to confront such a deeply philosophical and existential dilemma.

I have a long personal history with the topic of consciousness, shaped in part by my experiences with various mental health challenges. Combined with a lifelong obsession with the nature of existence, this has made the subject especially meaningful to me. As a result, this blog post will be far more personal than my previous entries.

Therefore, I should forewarn readers: this will not be a rigorously researched article like some of my others. Instead, it will be a stream-of-consciousness reflection on human awareness, AI 'pseudo'-consciousness, and the line that separates the two.

Scientific Consciousness

I would like to explore the scientific definition of consciousness and contrast it with the case study of an AI capable of presenting as a ‘conscious being’. Firstly, it's important to note that no widely accepted, definitive scientific explanation for consciousness exists. However, we can make educated assumptions based on human biology. To severely oversimplify a complex concept: biologically, consciousness is understood to emerge from complex patterns of neuronal activity and integration within the brain, giving rise to subjective experience and self-awareness. In doing so, the biological organism, wired in a certain way to induce this phenomenon, can be said to be ‘conscious’.

When contrasting this with artificial intelligence, it occurs to me that an AI could be developed in a similar way. If consciousness is born from biology, why could a mechanical ‘brain’ not be constructed the same way? If one were to be reductionist, humans are biological machinery. It should be possible to imitate the biological function of the human brain to create an artificial consciousness.

Currently, we are at the beginning to intermediate stages of neural network development in AI. It is not a leap in logic to suggest that in the future we could create a neural network capable of attaining artificial consciousness. That said, given that we currently do not fully understand human consciousness, I personally believe this would not be a deliberately engineered outcome. Instead, it feels more likely that true artificial consciousness, if it occurs, would be an accidental emergence of an AI system undergoing extensive self-learning and developing its own complex, unforeseen neural pathways. This is especially true if we get to a point where AI can self-modify, replicate components of its architecture, and continue learning without direct human supervision, creating its own neural paths and developing its own weights and measures for each individual node of the neural network. Even so, given our limited understanding of what true ‘consciousness’ is, it would be nearly impossible to determine whether any AI system is genuinely conscious or simply imitating it convincingly

(Formatting of this post was assisted with AI)