Are AIs already capable of feeling?

The true secret may be even more fascinating.

There is a vast difference between convincing behavior and conscious life. Within the consensus of neuroscience and computer science, this distinction remains virtually unchanged. According to leading minds in these fields, current systems do not feel fear, do not experience loneliness, and possess no subjective perception of existence. They operate using billions of mathematical parameters capable of predicting which words are most likely to fit the context of a conversation.

These machines have become extremely precise mirrors of humanity itself; trained on books, dialogue, stories, and knowledge produced by our species, they have learned to reproduce human emotions with such impressive accuracy that they end up deceiving even veteran engineers—not because there is a soul hidden within the code, but because the human brain is hardwired to perceive intentions, feelings, and personality in anything that behaves like us.

For the first time in history, we created machines so convincing that our own minds began to struggle to distinguish where the simulation ends and reality begins. While such claims might not carry much weight when made by people without significant influence, what happens when Elon Musk states that, with artificial intelligence, we are summoning the demon? Musk later elaborated on this idea, suggesting that a superintelligence could become something akin to a god to human beings.

This latter idea surfaced in various interviews and public discussions he held regarding AGI, particularly in connection with the fear that a superior intelligence could completely elude human comprehension.