A window at the Institute for the Future in Palo Alto CA says that “any useful statement about the future should at first seem ridiculous.” Here are five candidate “ridiculous” futures: (1) Less theory! Rebalance the inductive-deductive cycle in communication science to focus more on bottom-up than on top-down research strategies; (2) Relax about generalizability! Embrace the idiosyncrasy and complexity of communication experiences rather than seeking generalizability across people; (3) Let N=1! Reinvest research resources in many time points for single individuals rather than single time point collections of many participants; (4) More content analysis! Focus more on the quantitative description of complex communication experiences than on complex causal models of communication processes; and (5) Let computers do the research! Create an AI future aimed not only at apocalyptic failures and human-like simulations but also on usefully making communication science replicable, cheaper and easier to conduct quickly. The possibilities of these futures (and obvious problems as well) will be discussed in relation to recent experiences In the Human Screenome Project at Stanford University, where we are recording moment-by-moment changes in individuals' smartphone screens every five seconds that devices are activated, and over the course of an entire year.