Abstract: The development of artificial intelligence has been deeply entangled with biological thought. From a philosophy-of-biology perspective, we reexamine how the logic of life has been borrowed, transformed, and ultimately obscured in AI’s formation. The evolution from the McCulloch–Pitts network to deep learning reveals that early neural networks, though inspired by biology, gradually diverged from physiological mechanisms. The later influence of biological cybernetics revived attention to living system’s adaptability and self-organization, while Minsky’s pragmatic engineering approach attempted reconciliation but remained limited to a brain-centered view of intelligence. We argue that revisiting AI through existing forms of biological intelligence can overcome such “cerebrocentrism” and offer a new conceptual framework for understanding intelligence across artificial and biological domains.