Researchers at MIT introduced BoltzGen, a generative AI model designed to create candidate molecules for hard‑to‑treat diseases, and presented early demonstrations at a crowded seminar that drew attendees from academia and industry. The model integrates learned chemical priors with synthesis‑aware constraints to propose plausible, novel small molecules for downstream screening. Industry analysts and venture reports suggest AI advances are a key lever for reigniting venture activity in biotech, with PitchBook noting that AI‑driven platforms can de‑risk early discovery and attract capital back to early‑stage rounds. Investors are increasingly pairing ML teams with wet‑lab capabilities to validate AI outputs faster. Practical adoption hinges on assay integration, IP strategy, and reproducible translation of in silico proposals to validated hits — but the BoltzGen release illustrates growing momentum for AI‑first discovery pipelines.