We released the protein structures to the world knowing that the scientific community would find countless applications for them, far beyond what we could imagine.

We couldn't have achieved this without the foundational work of the structural biology community, which provided the initial 150,000 protein structures that our system learned from.

In just a few years, our small team was able to determine the structures of 200 million proteins, compared to the 150,000 structures revealed by tens of thousands of scientists over 50 years.

The launch of the AlphaFold server is crucial for providing the scientific community with access to our models, facilitating non-commercial academic research.

It's so awesome to meet developers from around the world and share the latest from Google DeepMind, especially seeing the excitement in the Gemini community, which is vibrant right now with many developers building with our models and asking for more features.

AlphaFold 3 allows our drug designers to see how their molecule designs interact with protein targets instantly, which is a game changer in drug design.

AlphaFold 3 represents a significant breakthrough in understanding biomolecular structure, allowing for experimental-level accuracy in predicting protein structures and their interactions with small molecules, DNA, and RNA.

AlphaFold has become a standard tool for biology and medical research, with over 2.5 million researchers worldwide using it in their critical work, significantly impacting the scientific community.

More than 2 million researchers worldwide now use AlphaFold, and it has had a significant impact on biology and medical research, but this is just the beginning of AI's potential in science.