LLMs, though limited, may spark testable hypotheses that inspire the creation of scientifically grounded, functionally intelligent systems.