Apodex Discovery is a groundbreaking framework for building and evaluating discoverative artificial intelligence. It comprises a foundation model, tools, control policies, and verification mechanisms to pursue extended, stateful, and verifiable investigations. The core components include a problem-scouting process, a common environment-task-episode abstraction, and the HDS6 evaluation metrics. The problem-scouting process surveyed 561 industries and collected 423 high-value real-world problems, selecting 20 for the initial release. The common environment-task-episode abstraction provides a unified interface for data, tools, constraints, feedback, trajectory recording, and verification of intermediate artifacts and final submissions. The HDS6 evaluation metrics assess tools, repair, alternatives, coherence, evidence, and scope independently of final-task success. Apodex Discovery achieved state-of-the-art results in AAV capsid design and drug repurposing and reformulation tasks. Blogger's Review: Apodex Discovery paves a new way for the evaluation and development of artificial intelligence, providing a more realistic and comprehensive assessment of AI models' capabilities and potential by simulating real-world challenges and problems.