作者: Joel Lehman , Kenneth O. Stanley
DOI: 10.1162/ARTL_A_00150
关键词: Biological theory 、 Artificial life 、 Natural (music) 、 Artificial intelligence 、 Relevance (information retrieval) 、 Cognitive science 、 Biology
摘要: An important goal in both artificial life and biology is uncovering the most general principles underlying life, which might catalyze our understanding of engineering lifelike machines. While many such have been hypothesized, conclusively testing them difficult because on Earth provides only a singular example from to infer. To circumvent this limitation, article formalizes an approach called radical reimplementation. The idea investigate abstract biological hypothesis by intentionally reimplementing its main diverge maximally existing natural examples. If reimplementation successfully exhibits properties resembling biology, it may support better than alternative inspired more directly nature. thereby principled common tradition defending minimizing deviations nature life. This work reviews examples that can be interpreted through lens yield potential insights into despite having purposely unnatural experimental setups. In way, help renew relevance computational systems for investigating theory act as practical philosophical tool separate fundamental features terrestrial epiphenomenal.