Since the first fiberoptic instruments, gastrointestinal endoscopy has shaped the field of gastroenterology and is now a key diagnostic and therapeutic tool. Compared with the initial fiberoptic endoscopes state-of-the-art optical chips (or charge-coupled device technology) allowed a quantum leap in image quality. Despite these advances, gastrointestinal endoscopy is far from being perfect. The diagnostic yield (eg, for adenoma detection rates) is highly operator dependent and there is still the need for sedation or even anesthesia to address discomfort during the procedure. Despite highly standardized cleaning and high-level disinfection the reuse of contemporary (and difficult to clean) endoscopes with multiple channels exposes patients to the risk of transmission of infections. Artificial intelligence and pattern recognition should eliminate interindividual variability including polyp detection rates, self-propelled, and (potentially remotely controlled) scopes with a soft shaft could reduce the discomfort during procedures and abolish the need for sedation and anesthesia altogether and single-use designs should eliminate the risk of patient-to-patient transmission of infections. While these innovations are feasible and could be implemented rapidly utilizing available technology, they require a paradigm shift affecting all levels of the value chain from the supplier of the instruments to the end-users. Some may negate the need for a paradigm shift, but it is evident that a major redesign of the endoscopic equipment is overdue to fully utilize novel technologies and most importantly ensure the best possible outcomes for patients.