Bart de Witte is chair of the faculty of future medicine at the recently founded futur/io, a new European education and research institute focused on exponential technologies and desirable futures. He is also a faculty member specializing in digital health at Switzerland’s University of Applied Sciences in Business Administration Zurich and the University of Applied Science in Burgenland, Austria. De Witte, who has over 20 years of global digital health experience in 26 countries, is currently director of digital health at IBM, focusing on digital health transformation, stimulating innovation and leveraging the company’s investments around artificial intelligence. He has been involved in leading-edge technologies and new IT businesses, including acting as a mentor in the formation and growth of a dozen startups within the healthcare sector. He recently spoke to The Innovator about the barriers to innovation in healthcare, the digital transformation of the sector and how AI could lead to a universal digital healthcare system.
Q : How is health care changing ?
BDW: Heath care is one of the industries that is really challenged. The model is most countries is not sustainable, the U.S. being a primary example. There has been a lot of discussion for 20 years about applying tech in the health sector but none of these technologies were really transformative, they helped organizations to become more efficient, but most technology was designed around systems and not around humans. This has led to the fact that today’s doctors spent about 50% of doctor’s time on their keyboards, and nurses use 70% of their time to document, administrate, communicate and coordinate, while they should be spending most of their time on patients. Howeer, due to the investments in digital systems, this technology has now created a lot of usable data and that is where the opportunity is. We have tons of healthcare data. If we apply AI methodologies such as Machine Learning or Deep Learning to health data, it will be really transformative. It is going to transform the core of medicine because the core of medicine is based on knowledge — knowledge that doctors use to take decisions to prevent, detect and treat diseases.
Q: What will this mean for the healthcare industry?
BDW: AI will hugely commoditize every kind of therapeutics from diagnostics to therapy and even surgery. In fact, AI experts predict that in 2053, 30 years from now, robots will outperform humans in surgery. The tendency is to describe this as a trillion dollar industry and a big opportunity. But there is one risk here for AI, the risk of centralized monopolies, something that we have seen within the search engine industry, as healthcare, like search engines, will be algorithm-based and run on platforms. These platform become more intelligent when they scale, the more they are fed by data the smarter they become. The so called feedback-loops are necessary for the algorithms to learn from the applied suggestions. When you search something on a search engine and you click on it Google knows your search was correct and if you have 100 million of these the more data and the more users you have the more search will improve. Transfer that model to healthcare and the company that has the most traction on this platform, ie the highest algorithm accuracy, the one that gathers the most data — they can build a monopoly. The risk here is that is that the medical knowledge that will be digitalized will be centralized, monopolized, privatized and become exclusive. So the question is do we keep healthcare as a public good or make healthcare knowledge part of a private industry?
Q: What would it mean to make this a public good?
BDW: For me the smartest thing we could do as a society, is to open source healthcare and create strategies to completely liberate data without having privacy issues. If we can create platforms that are available in a not-for- profit format and deliver these algorithms there is a huge chance we can create a universal digital healthcare system. We have never been so close to be able to deliver this. If you make it a public good and open source and make it a commodity — and I think we need to try to do this — if you are able to leverage communities that are working on health care developments and you have thousands of data scientists jointly in a non-commercial way similar to Apache or Linux open source platforms you could have the whole community developing for a purpose.
Q: How would such a system get started?
I have been setting up such a system which is due to launch in November as a moonshot project for future/io. The idea is to launch a platform that is looking to create universal digital healthcare by bringing a community together to solve specific challenges. We will start by applying AI to radiology and pathology and create a global network with meetups and events and gather together and develop algorithms and make them available in open source based on the purpose of creating universal health care with openness and transparency on the data and the algorithms used and validation of the algorithms by medical experts.We need to approach health care as a not-for-profit model because otherwise there is a very high risk that there will not be inclusive healthcare for all.
We in Europe have managed to develop healthcare systems that allowed us to have prosperity and longevity in an inclusive way. Thanks to technology life expectancy is accelerating pretty fast. If we don’t open source that knowledge it could be monopolized and become an exclusive good. If access to healthcare and thereby longevity is not equally spread, this will lead to disruptions in society. Already you see this in the U.S. where the life span has been declining and poor people do not have access to the health and social services they need. Healthy living and healthy aging are an exclusive good, leading to inequalities in society. I strongly believe that as a technologist we have a responsibility to create a future that is desirable forgenerations to come. If AI has the possibility to commoditize healthcare knowledge and making if publicly available, then this is a thing we have to strive for.
Q: How do you think the established health care industry will react?
Not all will like this initially. But as we have seen with other open source movements, there is plenty of opportunity for businesses to create great products and services with open sourced code. The author Thomas Friedman wrote in his book The World is Flat about how IBM understood that when they spend millions in the open source community it would help them to sell more web servers and become more competitive towards Microsoft.
Q So what is your advice to traditional companies in this space?
Be open. When you develop AI strategies try to embrace the openness and share. Give back to the community and at the same time leverage open source and add true value.