Overview
- TikTok chief executive Shou Zi Chew publicly said he expects major AI-enabled discoveries in fields such as physics, biotech, chemistry and materials science within two to three years, remarks reported from a Singapore event.
- ByteDance has been expanding beyond consumer features into scientific AI, releasing multimodal Seed2.0 and Seed2.1 models and unveiling Protenix-v2, an AI system for biomolecular structure prediction and design.
- A recently announced $1.8 billion public–private initiative involving the U.S. government, Meta and Google will fund models that predict how cells behave to speed biological research and drug discovery.
- Chew warned that more powerful, agentic AI raises cybersecurity risks because users may grant broad permissions to autonomous systems and bad actors can use models to scale fraud and spam.
- The cluster of near-term forecasts heightens reputational and governance stakes for companies and regulators, and it makes scientific validation, model alignment and secure data practices the key things to watch next.