In a fairly short period of time, artificial intelligence has become a regular facet of our everyday lives, helping us get through each day easier, faster and smarter. We use AI as we navigate our way through town by calculating where traffic or construction lies in wait, it curates playlists that cater to our particular tastes in music, and alerts us to banking fraud or intruders in our home.

But what if AI could be used to help combat threats to human health, especially in hard-to-reach regions? Diseases like Ebola spread quickly, and health workers often face unique infrastructure challenges due to access or instability. On top of that, doctors on the ground lack access to critical data that could help stop the disease from spreading. How could AI help fill these data gaps, and what data is actually useful? Sumi Paranjape, Vulcan’s Director for Health and Science on the Philanthropic Impact Team, tackled these very questions and more at the SXSW Interactive panel discussion “The Role of AI in Reducing the Impact of Pandemics.”

Sumi at SXSW Interactive

We caught up with her to talk about some of the highlights and key takeaways from that discussion.

How has Vulcan historically been involved in industries like AI and global health?

At Vulcan, we use data and technology to drive our efforts and ensure a positive and sustainable impact. We strive to leverage tomorrow’s technology to solve some of the world’s biggest challenges today.

Our founder, Paul Allen, was a technologist. He believed in the power of data to save and improve lives. He founded the Allen Institute for Artificial Intelligence in 2013 and has pursued projects in “AI for the common good”, tools like Mosaic and Semantic Scholar.

When the West Africa Ebola outbreak exploded in 2014, Paul quickly committed $100 million to help catalyze progress in technology, diagnostics, data and many other innovative areas. In the case of that Ebola outbreak, very little of the $100 million was granted to anything related to AI because at that time there weren’t many real-world applications that were ready for use in an emergency setting, but fast forward to 2019, and we’re facing a very different Ebola outbreak in the northeastern Democratic Republic of Congo in North Kivu and Ituri provinces. Now we can ask bigger questions by pulling in newly available data layers and creating real-time, geographically targeted insights.

What is your vision for how AI can stop pandemics?

AI isn’t a silver bullet for pandemics, but it is a powerful tool that can help generate information that enables more precise and effective strategies for prevention, detection and response to outbreaks. It enables health workers to save lives faster.

Ending all infections, whether it’s the common cold, influenza, measles or Ebola will be very difficult, but by harnessing the power of data and advanced analytics, we can try to understand when the first cases are happening and make sure we control the spread quickly. Doing this, we can end dangerous infections before they become outbreaks or sustained, far-reaching pandemics.

How exactly can AI be applied to pandemic response?

AI can help with many things from establishing priorities for limited funding and resource allocation to improving care delivery, to prediction and disease monitoring. Right now, there are many areas where AI and advanced analytics can help us. In particular, we can leverage data to generate insights that will help us respond and contain outbreaks quickly.

There is data that can be used to help understand when an outbreak is starting, and AI can help us develop strategies for intervention during a response to an outbreak. AI can also help improve the efficacy of other tools in place by informing vaccination strategies, helping to predict where outbreaks may spread and even helping to develop insights that will enable us to rebuild the trust of affected communities.

Even in this current Ebola outbreak, we can improve our ability to respond to the outbreak by using mobility data to predict where Ebola cases could surface next or to identify the factors that have Ebola persisting in one area rather than others.

What challenges does AI face in this space?

Data privacy issues are challenging. We must ensure that we protect people’s privacy but still find ways to be able to aggregate data so that we can generate information that we can use for maximal societal benefit. New advances in cancer treatment are a good example of what we can learn when we share data while protecting patient privacy.

We also need to be able to leverage different sources of data. For prediction, we’ll need to leverage ecological data, weather data, data on the pathogen (virus or bacteria), historical data, data on healthcare, etc. This means we need an interdisciplinary approach. Perhaps the biggest challenge to effectively deploying AI is making sure that the insights are actionable.

One of the most important considerations relates to the fact that, in order to generate actionable information, we need to involve stakeholders across several disciplines and organizations. Only then can we be sure that we are leveraging AI for maximum impact.

At Vulcan, we’re very excited to be exploring this space because we believe there is an important role for a philanthropic organization who can convene multiple stakeholders and who is committed to developing sustainable and scalable products. We measure the impact they have on ending the threat and intensity of pandemics when we’re tracking success.

What is the Ebola Response Accelerator Challenge?

The Paul G. Allen Family Foundation just launched the Ebola Response Accelerator Challenge to help catalyze and quickly implement innovative solutions for the current Ebola crisis in the DRC. People are dying, we have to act fast.

We’re focused on solutions that will overcome some of the most daunting challenges specific to this outbreak – issues of community distrust, communication breakdowns, and safety challenges that prevent patients from accessing treatment. One of the priority areas we’re looking for from proposals in is accelerating and improving data, analysis, and communication. We hope to hear from people thinking of innovative concepts that can help us support the frontline presence, can be deployed immediately to address the current outbreak, and can be aligned with the current ongoing strategic response. Head to the Challenge page to learn more about submitting a proposal.

 

Technology & Science Africa disease Ebola innovation pandemics technology

READING LIST

The Role of Artificial Intelligence in Reducing the Impact of Pandemics

By The Editors

Allen Institute for Artificial Intelligence to Pursue Common Sense for AI

By The Editors

How Artificial Intelligence Could Save the Planet

By Lucas Joppa

1 Year, 10 Innovations From UW's Paul G. Allen School That's Making the World a Better Place

By The Editors

Building Partnerships to Reverse Seattle's Homelessness Crisis

By Paul Butler

Ebola Outbreak in West Africa: Now Is the Time to Help

By Paul G. Allen