And here is something that all technology should be, working for humanity.
Google Deep Mind in collaboration with Yale University,
presented an advance that could radically change the way we fight cancer, a new artificial intelligence model called C2S-Scale 27B was able to predict a combination of drugs capable of,
“awaken the immune system” and make it recognize tumors previously invisible to the human body.
The most surprising thing is that this hypothesis did not come from a scientist, it arose from AI itself and was confirmed in living laboratory cells. The model based on Google's Gemma architecture is one of the most powerful systems ever created for understanding cellular behavior. He was trained to “read the language of cells,” interpreting how they communicate and react to medications under different biological conditions.
The AI analyzed more than 4,000 compounds used in patients and tumor samples, virtually simulating thousands of cellular contexts until it found something that could convert “cold” tumors, those that go unnoticed by the immune system, into “hot” tumors, that is, visible and attackable by the body's natural defenses, which is where the discovery arose.
The exact combination to locate and attack cancer.
The C2S-Scale 27B identified that the combination between the drug CX4945 and interferon at low doses could generate a significantly greater immune response than either of the two alone. Isolates produced almost no effects together, however, the result was a 50% increase in antigen presentation, allowing the immune system to recognize and react against the tumor.
Tests carried out on human neuroendocrine cells that were not part of the model's training data confirmed the AI's prediction, demonstrating that the system not only analyzed data, but also reasoned within a biological context, something that until now was considered exclusive to human intuition.
This combination of contextual reasoning and laboratory validation represents an unprecedented leap in oncology research, suggesting that AI models can act as virtual laboratories, performing millions of theoretical experiments before any physical testing.
With this, Google Deep Mind opens a new era for science, that of artificial intelligences capable of formulating complex biological hypotheses and revealing paths that humans alone may never discover. Yale researchers are now studying the cellular mechanism behind this synergy between the two drugs and new predictions are already being tested.
If the results are confirmed in clinical studies, this AI could transform the way we design treatments and perhaps change the fate of millions of patients. It will be that in the future the cure for cancer will not come from a human laboratory, but from the logical mind of an artificial intelligence.
