We are in an era where applied science, human creativity, and new technologies provide deep insights into humanity's fundamental questions. While scientific progress is often seen as relentless, it has slowed over recent decades. However, recent advances in fields like artificial intelligence (AI) and quantum computing have dramatically accelerated discovery in various areas, from healthcare to environmental solutions. These advances result from decades of collaboration among researchers, technologists, policymakers, and civil organizations, creating a blueprint for how AI can improve human life. Today, The Royal Society and Google DeepMind are cohosting the first AI for Science Forum in London, bringing together the scientific community to explore AI's transformative potential in science and the role of public-private partnerships in innovation. Some notable recent achievements include: 1. **Protein Prediction**: Google DeepMind's AlphaFold 2 model has predicted the structures of 200 million proteins, accelerating progress in medicine and tackling environmental issues. The follow-up model, AlphaFold 3, builds on this success. 2. **Brain Mapping**: In collaboration with Harvard, Google has mapped a tiny section of the human brain in unprecedented detail, releasing the dataset to advance health research. 3. **Flood Forecasting**: Google's flood forecasting AI, started in 2018, now predicts floods up to seven days in advance in 100 countries, serving 700 million people. 4.

**Wildfire Detection**: Google partnered with the U. S. Forest Service to develop FireSat, an AI and satellite system that detects wildfires quickly, aiding rapid response. 5. **Weather Prediction**: Google DeepMind's GraphCast predicts weather and cyclone tracks up to 10 days in advance with greater accuracy than traditional models. 6. **Mathematical Reasoning**: Google DeepMind's AlphaGeometry and AlphaProof models solve complex geometry problems, advancing AI's reasoning capabilities. 7. **Quantum Chemistry**: In collaboration with UC Berkeley and Columbia University, Google conducted large quantum chemistry simulations, enabling precise chemical predictions. 8. **Materials Science**: Google DeepMind's GNoME tool has identified 380, 000 stable materials, potentially leading to advances in solar cells, batteries, and superconductors. 9. **Nuclear Fusion**: Google DeepMind developed AI that autonomously controls plasma in fusion reactors, a step toward achieving stable fusion for clean energy. These moments mark significant progress in the scientific landscape and highlight the potential for continued innovations.