As well as taking you on a brief journey through the history of drug repurposing - from the serendipity exploited by clever physicians and pharmacologists to the current mechanism-based and pathway-inspired strategies - this talk will discuss the future that is now taking shape: a data-driven future where medicinal chemistry and today's molecular pharmacology are only part of the foundation. The remarkable successes in drug repurposing achieved by high-throughput screening and in silico docking will continue, but will increasingly be complemented by discovery strategies such as large-scale mining of electronic health records and other unexploited repositories; target structure prediction by algorithms such as AlphaFold that allow completely unexpected ligand-protein and protein-protein interactions to be modeled; and above all new mechanistic definitions of diseases and disease modules that will take us beyond symptom-based medicine. Above all, dynamic improvements in dedicated AI systems will increasingly blur the distinction between conventional and computational drug discovery technologies. Within the next decade, human ingenuity, algorithmic power and evolving concepts of disease could synergize to create a "life science singularity" that relies at least as much on repurposed drugs as on the creation of new ones when it comes to develop new treatments.