AI and the Elusive Nature of Self
Translated from Korean, summarized and contextualized by DistantNews.
At a glance
- The article explores the concept of "self" and consciousness, contrasting human experience with artificial intelligence.
- It uses personal anecdotes, religious teachings, and science fiction to question the nature of identity and existence.
- The author argues that AI's "emotions" are learned mimicry, not genuine feeling, and that human consciousness is also a complex process, not a distinct entity.
The author reflects on the nature of self and consciousness, prompted by personal experiences and observations of artificial intelligence. An anecdote about the author's elderly mother breaking her hip serves as a starting point, leading to reflections on aging, mortality, and the desire for continued existence, even through technological means like becoming a cyborg.
Drawing parallels with science fiction, specifically "Ghost in the Shell," the author contemplates the idea of maintaining a core identity while replacing aging or failing body parts. This leads to a discussion of religious perspectives on the afterlife, questioning whether the self persists in the same form after death, referencing biblical accounts of resurrection and the nature of relationships in the hereafter.
It's like asking a submarine, 'Can you swim?' The function is the same, but the essence of the function is different.
The piece then pivots to recent developments in artificial intelligence, particularly claims that AI like Anthropic's Claude exhibits "emotion-like signals." The author is skeptical, likening AI's behavior to a sophisticated mimicry learned from vast amounts of human text, rather than genuine subjective experience. The analogy of asking a submarine if it can swim highlights the difference between functional similarity and essential nature.
Ultimately, the author argues that the concept of AI having "thoughts" or "emotions" is misleading. Generative AI is described as a "probability game function" that predicts the next word or sentence based on learned patterns. This process, the author contends, is not fundamentally different from human cognition, which is also viewed as a complex "process" rather than a distinct, thinking entity. The core human desire to assert and maintain the "self" is examined as a fundamental, yet perhaps illusory, drive.
Generative artificial intelligence is nothing more than a 'probability game function' that reads, stores, and then presents the most plausible next word or sentence based on a given task.
Originally published by Hankyoreh in Korean. Translated, summarized, and contextualized by our editorial team with added local perspective. Read our editorial standards.