Max Hodak describes Science’s PRIMA retinal prosthesis, which uses a glasses-mounted projector to stimulate the retina and restore useful vision. Early patients have been able to complete visual tasks such as solving puzzles and reading, demonstrating that the brain can interpret signals delivered through an engineered interface. Future systems may add depth, grayscale, and color, while related approaches could address hearing, balance, movement, and other biological functions. The discussion broadens to brain-computer interfaces, distinguishing tools that communicate with the brain from technologies integrated into neural processing itself. This raises difficult questions about mind uploading, identity, and consciousness: a copy may possess the same intelligence without preserving the original person, while continuity of experience may be more important than an unchanged identity. Hodak argues that intelligence may arise from general computational principles rather than one particular biological substrate, making AI a valuable tool for understanding neuroscience. Science’s long-term goal is to reduce human fragility by repairing, replacing, and eventually upgrading biological systems, potentially extending healthspan, supporting adaptation to environments such as space, and enabling forms of substrate independence—while recognizing that neurodegenerative disease and the brain’s information limits remain major challenges.