How AI and human intelligence will defeat cancer

In 2016, a significant milestone for mankind was accomplished: artificial intelligence (AI) beat the world champion at the game of Go. For context: Go is a board game that was previously thought to require too much human intuition for a computer to be successful, and so it was a North Star for AI.

For years, researchers tried and failed to create an AI system that could beat humans at the game. Until AlphaGo.

In 2016, AlphaGo, an AI system developed by Google’s DeepMind, not only beat its human counterpart (Lee Sedol); It showed that machines could find game strategies that no human would. AlphaGo shocked the world when it performed its unimaginable #37 move. It was a move so counterintuitive and strange to human experts that after AlphaGo played it, it stunned and confused Lee and all viewers and world experts. It ultimately led to the triumph of technology during this game.

AI vs. Cancer: In Search of Move 37

In addition to illustrating the potential of AI in this context, the Go game demonstrated that AI could and should help humanity develop the Move 37 for significant real-world problems. This includes fighting cancer.

As with board games, there is a special element of a game in the proverbial “competition” between the human immune system and cancer. If the immune system is the cop guarding the body’s health, cancer is like a gangster trying to evade capture. While the “immune system police” search for harmful cancer cells, viruses, infections and any diseases, the cancer develops various tactics of subversion, deception and destruction.

Let data expand our intuition

Centuries ago, scientists and physicians largely groped in the dark when it came to curing diseases, relying solely on their intuition. Today, however, humanity is uniquely positioned to take full advantage of available resources with advances in high-throughput and biological data measurement. We can now create AI models and use any available data to allow these AIs to augment our innate intuition.

To illustrate this concept more clearly, consider the case of CAR-T cells that have been edited with CRISPR (a gene editing technology) to create a promising therapeutic option to treat cancer. Many current and previous approaches in this area have relied on the intuition of an individual researcher or academic group to prioritize which genes to test. For example, some of the world’s leading experts in genetically engineered T cells had the idea of ​​trying to turn off PD1, which failed to improve patient outcomes. In this case, the genes were not compared directly, and a great deal of human intuition was required to decide how best to proceed.

Recently, with advances in high-throughput single-cell CRISPR sequencing methods, we are approaching the possibility of simply testing all genes simultaneously, equitably, and in different experimental scenarios. This makes the data more suitable for AI, and in this case we have the opportunity to have AI help us decide which genes are the most promising to modify in patients to fight their cancer.

The ability to run large-scale AI experiments and generate data to fight cancer is a game changer. Biology and disease are so complex that current and past strategies, largely driven by human intuition, are unlikely to be the best approaches. In fact, we predict that within the next 10 years we will have the equivalent of a Move 37 for cancer: a therapy that may seem counterintuitive at first (and which human intuition alone would not lead to), but which ends up shocking us all and wins the game for the patients.

Luis Voloch is CTO and co-founder of Immunai.