Daphne Koller is bilingual not only in Hebrew and English, but also in the research languages of machine learning and biology. As a result, she views human biology through the lens of data.
Machine learning and its ability to process information beyond human capabilities allow it to tackle an array of important problems. Machine learning’s key drivers are massive data, powerful compute capabilities, and better models.
In the biological sciences, there have been amazing advances in medicine, from polio vaccines in 1954 through biologics to treat autoimmune diseases, cancer immunotherapies, cystic fibrosis treatment, and today’s cell therapies. Still, however, drug discovery is an time-consuming, hugely expensive undertaking, making it ripe for disruption.
Advances in genetic and other biological data generation—including creation of cellular systems, perturbation of those systems using fine-grained molecular scissors, and high-content phenotyping—lay the groundwork for application of a quote by journalist Thomas Friedman “Big breakthroughs happen when what is suddenly possible meets what is desperately necessary.”
Creating a single framework that merges biology and data—until now separately evolving areas of technology—is the next step. It will enable bio-data factories, or factories for the creation of biological data that drives machine learning, able to revolutionize predictions on disease, biology, and drug development.