Emerging evidence suggests that the introns and intergenic sequences of the genomes of higher eukaryotes (the ``junk'' DNA) codes for a vast, RNA-based, genetic regulatory network. It is believed that this network is responsible for the variety and complexity of terrestrial life. We conjecture that this regulatory network is more properly viewed as an RNA ``community'', composed of a rich and largely unexplored biochemical web of RNA interactions. Viewed as an RNA-community, we hypothesize that the RNA regulatory network of higher eukaryotes can re-wire itself, and employ various and evolvable mutational strategies in response to external pressures. Thus, we argue that much evolutionary change is due to intracellular, RNA-mediated learning processes. Successful strategies and pathways are then recorded (hard-wired) into the DNA genome via reverse transcriptase. We present evidence which is consistent with this viewpoint, and make specific predictions which could be used to test the utility of our framework. If essentially correct, the RNA-community view of eukaryotic cells could reconcile measured point mutation and gene duplication rates with actual rates of evolutionary change. Futhermore, the RNA-community view of eukaryotic cells suggests that agent-based modeling techniques, used in mathematical economics, game theory, and neuroscience, will likely be as useful in understanding the functioning of eukaryotic cells as the pathway-based approaches of systems biology. We conclude this paper by arguing that a sufficient amount of biological knowledge has been accumulated to initiate a systematic program of experimental and computational studies of the origins and macroevolution of terrestrial life.