From YouTube to Fitbit, Google Health’s annual event last week showcased a bevy of health-focused improvements the company’s infusing into various products and services. Greg Corrado who leads Google’s Health AI division helped me unpack Google’s approach towards health and user well-being, starting with the device nearly all of us can’t think of living without – the humble smartphone.
A majority of present day smartphones are far more sophisticated than we think. Greg Corrado explained how his research at Google involves trying to use smartphone cameras for detecting diabetes and sensing irregular heart sounds with smartphone mics and much more, all because of the fundamental promise of the smartphone.
“What brought the smartphone to prominence is its usefulness as a communication device, as a way for people to communicate with each other, and fundamentally the provision of healthcare is also about communication, sensing, and measuring. I think that there’s a possibility here for smartphones to start to play some of the roles that other pieces of technology that were previously healthcare-specific might have played, both in the hands of doctors and caregivers, and also for people who are caring for themselves,” said Greg Corrado.
He further suggested how people already get a lot of information that’s related to healthcare from their phones. Beyond making doctors appointments through their smartphones, Google’s seriously pursuing the exploration of smartphones as almost an early diagnostic and measurement devices, specifically with respect to their camera sensors and microphones, he added.
“Apart from their demonstrated communication utility, it seems like a worthwhile endeavour to understand how smartphones can help us bring care to more people around the world. If smartphones have made it to where people are and where people need help, I definitely want to see if we can build upon Google’s platforms to make it (provide health-related improvements) available at scale. And we are trying to build these healthcare pillars through our products and services in the most open way possible, so that it’s as useful to as many people as possible,” he emphasised.
As a Distinguished Scientist at Google Research, Greg Corrado joined Google to push the frontier of biologically inspired computing, and in 2011 co-founded the Google Brain Team, which has helped catalyse the broad adoption of deep neural networks across various industries. But overcoming health-related challenges for parts of the world that can truly benefit from Google’s initiatives remains Greg’s main focus.
“Sitting in Palo Alto, California, I think that a lot of folks here get reasonably good healthcare, but there is much more opportunity to help folks around the world. And that requires development in collaboration with people who are really working on the ground at local levels. That’s part of the reason why our partnerships in India, Thailand, and on the African continent, that’s a huge motivation for people like me here at Google right now,” Greg mentioned.
“I strongly believe that we need to engage with local communities, with local developers and engineers to be able to provide tools and demonstrations of technology and templates that make access to these technologies more available,” said Greg Corrado.
According to him, it’s not about a specific algorithm that Google ends up developing will become ubiquitous across the world, but it’s about giving a spark to young minds. “If young engineers working in India right now think, if Google can do this, what else is doable? And that’s when we will be able to truly democratise these health-related technological advancements and truly give people the power and opportunity to scale these things and apply to local problems. One of the things that I’ve really enjoyed in my time at Google is our willingness to publish scientific research, really explain how things work and to open source code and tools that we ourselves have used so that others can then use them to build things the way that they see them fit.”
Greg Corrado argues no single company has a monopoly on creativity and problem solving, and engineering advancements like these (particularly related to health and well-being) should be more ubiquitous and accessible to most of the world.
“Sometimes other organisations engage in sort of a focused approach to just generating novel IP or developing proprietary devices, believing that they themselves try to bring to market at scale,” said Greg, which he believes is the wrong approach to solve health-related problems.
“Our mission here at Google is about organising the world’s information and our core technical capabilities excel in distributed computing, in infusing artificial intelligence and making it more useful to people. I don’t believe that we are a traditional healthcare company, we will never become a traditional healthcare company. But I believe that through artificial intelligence, cloud computing and mobile computing, we can be of use to those who really are the medical specialists, community workers and professionals who work in the healthcare industry. And also to public health officials who are trying to make the best decisions for their country’s population overall. How can we be helpful in those contexts? That’s what we are working towards,” according to Greg Corrado.
As an academic, Greg Corrado is acutely aware of the autonomy of users when it comes to interacting with AI at scale. Systems need to be designed with that consideration in mind, he emphasised, where human choice needs to be paramount.
“We believe in respecting the user’s ability to make the right decisions. Because in the healthcare space, we have users across the spectrum from end consumers to trained medical professionals, and their needs are different – which we need to respect,” underscored Greg Corrado. “I view our role as providing tools and technologies that are fundamentally safe, because we don’t want to expose users to something that is unsafe or demonstrably inaccurate. For everything we do related to health and AI, we want it to be safe and accurate, quantify its degree of safety and accuracy, and then trust the user to make the appropriate decision based on their individual or expert judgement. That will always be our goal here at Google.”