Five years after the release of AlphaFold 2, Google DeepMind’s protein‑structure AI has moved from breakthrough to core scientific infrastructure, reshaping drug discovery and graduate training. DeepMind scientists including John Jumper and Pushmeet Kohli say AlphaFold and successor models are now taught as standard tools in molecular biology programs and are accelerating research timelines that used to require costly laboratory structure determination. The system predicts a protein’s 3D shape from its genetic sequence — a problem that historically demanded slow experimental methods — enabling universities and research centers to prioritize experiments, prototype therapeutics faster, and train students in AI‑assisted lab methods.