In the search for approaches to Artificial General Intelligence (AGI), I began to see an analogy with living organisms. In biology, DNA is not a static archive but an instruction to assemble a human being, a biochemical recipe executed molecule by molecule. Viruses need even less: only RNA or DNA, which on their own are not alive but 'come to life' when they enter a cell and access resources. By extension, AGI can be viewed as a digital organism with a minimal 'code of everything' a form of digital DNA. Outside its environment, this code is just a digital unit, but within computational infrastructures it unfolds into a full system. This code does not store all knowledge directly; it encodes rules for reaching ontologies, loading modules, and assembling intelligence. Tasks activate different 'genes,' similar to epigenetic switching in cells. This view raises practical questions: how AGI will consume resources, how it might replicate safely, and what digital ecology will sustain it.
Yevheniia Babenko (Fri,) studied this question.