Sundar Pichai, CEO of Google, reveals on X that an AI model has formulated a therapeutic hypothesis against cancer. This breakthrough is a result of collaboration with Yale University.
Artificial intelligence is joining forces with medicine — it’s hard to find a more noble cause. Researchers from Google DeepMind are collaborating with Yale University to make sense of C2S-Scale 27B. This model, based on Gemma, explores avenues that human researchers have yet to master. Algorithms venture where science still hesitates to better understand cancer.
C2S-Scale 27B, The Google Model in the Fight Against Cancer
In a detailed article, the DeepMind division unveils the workings of C2S-Scale 27B, a model featuring 27 billion parameters. Specialized in cellular and molecular biology, it differs from general AI by its medical focus. Its foundation, Gemma, serves as a robust base to explore the mechanisms of life.
The Google model meticulously analyzes each cell, tracking those that could reveal cancer’s secrets. C2S-Scale 27B detects invisible patterns in biological data and predicts unexpected molecular interactions.
The model identifies a vulnerability specific to cancer cells, a weakness that could slow their growth. Its hypothesis emerges from massive analysis based on millions of cellular sequences.
These discoveries then guide experts toward new therapeutic avenues. Teams subsequently test these ideas in the lab.
Experiments on living cells confirm the AI’s predictions. The cancer cells react exactly as the model anticipated.
Yale University, Guaranteeing Scientific Credibility
The collaboration with Yale University enhances the scope of the project. Experts in oncology oversee each stage, ensuring scientific rigor.
These researchers also integrate data from real laboratories to achieve optimal results. C2S-Scale 27B learns from these massive datasets, refines its hypotheses with each iteration, and focuses its power on biology.
Sundar Pichai applauds this advance as a new path in the fight against cancer. Researchers are already preparing for preclinical tests on animal models.
If results are promising, clinical trials will follow to assess the therapeutic potential of this approach.
