As technology advances and researchers and engineers are building artificial assistants, curing diseases with AI, and machines are learning their own languages, it’s forgivable to assume that humans are harnessing machines to become invincible. Whilst we can optimise business efficiency and even find cures to rare diseases in the future, there is one thing that remains unsolved. Living longer, and staying healthier.
Calico Labs, a research and development biotech company, is on a mission is to understand the biology of aging with the goal of combating aging and associated diseases. As we age our bodies begin to fail us, but why? Calico’s mission is to understand the biology behind this and develop interventions that enable people to lead longer and healthier lives.
At the Deep Learning Summit in San Francisco next January 25 & 26, Daphne Koller, Chief Computing Officer at Calico Labs, will share her work on the advanced research technologies that she is using to help Calico on their mission. This is no easy feat however, as the process ‘requires deep collaboration across multiple research and computing disciplines’, as well as a world-class team of computer and research scientists.
Sharing with us how she first got into working in AI, Daphne explained that she ‘started out being very interested in game theory and understanding how multiple agents reason in an uncertain and noisy environment’ which then led to machine learning and constructing models that could sift through noise and ‘help autonomous agents make good decisions’. This then led to an interest in the practical applications of these approaches to real-world problems in healthcare, leading to her involvement in biology and medicine. Teaching machines to make ‘good’ decisions requires huge amounts of data, and medicine has some of the richest data sets available.
Intrigued to hear more about her work ahead of the summit early next year, we spoke to Daphne to get some answers to some of our most pressing questions:
Today, high-throughput experiments are leading to massive amounts of biological data (e.g. genomes, proteomes, transcriptomes, electronic health records) that are produced at an unprecedented rate. We believe that computing, especially machine learning capabilities, will help us better analyze and understand these data to uncover novel insights about underlying biological pathways, and ultimately help us extract relevant signals that could be important in aging or age-related diseases.
At Calico, the team we are building are working side-by-side with colleagues in the biomedical sciences to design experiments, develop new analytical techniques that extend our computer science and machine learning methods and help us answer some of the most important questions about how we age.
My work in that space ranges from understanding the molecular pathways that underlie gene regulation to the analysis of breast cancer pathology images, a project that uncovered some features of the tumor that were entirely unexpected in being associated with survival.
Human biology is the single most complex system I’ve ever worked with, and how we age is one of the most fundamental questions of our time. We know that the single greatest risk factor for most non-pediatric diseases, including cancer, cardiovascular disease and neurological disorders such as Alzheimer’s, is age. Every year beyond the age of forty, our risk for these disease grows exponentially. It is important that we work to understand what’s driving this process, and to identify potential interventions that might help increase one’s health span, if not life span. At Calico, we focus significant time and resources toward the development of new technologies from multiple areas, such as genomics, proteomics, mass spectrometry, and microscopy, to provide new insights into this complex problem. From my own perspective, perhaps the most important technology will be computing - which is key to synthesizing the huge amount of data that we are now able to collect, and to deriving usable insights.
There are many aspects of scientific research and medical practice where AI and DL can play a critical role. During the next five years, and beyond, we’ll see the ability to sift through and draw informed conclusions from data in nearly every aspect of healthcare, such as: understanding key biological mechanisms from genomic data, helping design drugs that intervene at particular targets, tailoring treatment to a patient’s particular needs, and tracking the health of patients to allow earlier and more effective intervention. I believe that we are at an inflection point where our computing power and technology are running together and we will be able to interpret data to be able to make a real difference.
Medical professionals all have the goal of helping patients, and I believe that most are excited about the possibilities that AI can bring to improving patient care. My experience is that they have, and will continue to embrace new technology, including AI, if it provides better information that allows them to make more informed recommendations that affect patient care. Healthcare professionals are already using genomic data to help make decisions related to precision medicine and targeted therapies. They want the technology and tools to be validated, and to fit well into workflows that can best benefit patients. Additionally, medical professionals face the real challenge of trying to review and understand an enormous amount of information, including relevant peer-reviewed papers, being published every day, in order to make important medical recommendations for their patients. Levering AI techniques can help them by providing tools to help them synthesize the most important information they need.
One of the biggest challenges we face is that the huge influx of data that we can now access opens many opportunities, and we are limited in how many of these we can pursue. We have many exciting projects underway, but there are still others that might be just as valuable. To help with that, we are working on identifying and recruiting the best people who have machine learning/AI experience, a deep passion for science, and are open to challenging conventional wisdom, taking smart risks, and being part of a team that is trying to figure out one of life’s great mysteries…how we age.
|RE•WORK are returning to San Francisco for the largest Global Deep Learning Summit next January 25 & 26 where Daphne will be joined by other leading experts in AI. Historically the biggest summit of the year, the Deep Learning Summit will be joined by the Deep Learning for Enterprise Summit where we will explore applications of deep learning to optimize business efficiency, and the AI Assistants Summit that will showcase the opportunities of advancing trends in AI Assistants & their impact on business & society. |
Already confirmed to present their most cutting edge work are: