Network Spotlight: Kush Varshney and Aleksandra Mojsilović, IBM

October 31, 2017
Posted in
October 31, 2017 casper

Kush Varshney and Aleksandra Mojsilović at IBM’s T. J. Watson Research Center in Yorktown Heights, NY where they lead research on artificial intelligence and machine learning. (Photo: IBM)

What does the future of food look like through the lens of artificial intelligence (AI) and machine learning? These technologies give us unprecedented capabilities to find patterns in enormous datasets, and two data scientists in our network are looking into how these capabilities can inform efforts for social impact.

Kush R. Varshney is a research staff member and manager with IBM Research AI, working from the T. J. Watson Research Center in Yorktown Heights, NY. Aleksandra (Saška) Mojsilović is an IBM Fellow and a global lead for IBM AI Science Research. Together, they co-direct the IBM Science for Social Good initiative and conduct fundamental research on machine learning, statistical signal processing, and data mining.

With the Science for Social Good initiative, Mojsilović, Varshney, and their colleagues are tackling challenges ranging from energy conservation to combating the opioid crisis. In the food space, Science for Social Good is collaborating with the New York City-based emergency food provider St John’s Bread & Life to model and share best practices with other emergency food providers. Other food-related AI projects include Chef Watson, a Web app that draws on a vast trove of food data to generate pairing ideas and recipe suggestions. (Chef Watson’s recipes are creative, writes the Washington Post’s Caitlin Dewey: they include “[a] weird beef burrito accented with chocolate and edamame. A risotto studded with candied ginger, of all possible things [… and a] pumpkin-ricotta cheesecake with savory mushroom meringues.”)

LAUNCH Director of Engagement Davar Ardalan conducted this interview with Mojsilović and Varshney.

1) You are data scientists who have dedicated your work in the past few years at IBM research to science for social good. Tell us why and share a couple examples of how you’ve made an impact.

Kush: I believe that the most capable among us should take on the most difficult problems confronting humanity.  Our projects, described on the webpages of the IBM Science for Social Good initiative are small illustrative steps that partnerships between innovative NGOs and technologists like us can enable.

Saška: We started the program to raise awareness, and to show that data science, machine learning, and artificial intelligence could be critical to addressing some of the world’s biggest challenges. What we call Big Data today is really big — not in terms of volume, variety, velocity, and veracity — but in terms of how many questions we can answer and how many problems we can tackle. And these are some hard problems. They will make us think, and force us to be more innovative, as scientist, engineers, developers, and they will make our companies make better products..

2) For those not familiar with this emerging space. Define AI and machine learning as you see it.

Kush: Data includes numbers and attributes, but also text documents, audio, video, electrocardiograms, seismograms, and more.  Machine learning is a category of artificial intelligence technologies that allows computers to make sense of this data and allows computers to help people make sense of this data.

Saška: When we talk about machine learning and AI, people often ask us about the future, and where technology is heading. Where AI is heading. They ask if we are going to be over-consumed with intelligent machines. We tell them that technology is impartial, and it’s upon us to decide how to use. It’s up to us to decide what kind of applications we want to build. We have a responsibility as people and as scientists. By giving the Science for Social Good Fellowships to young researches in the field, we hope to influence new generations of scientists to think and act the same.

3) We recently met you to learn more about ways artificial intelligence and machine learning can disrupt the food and nutrition space. Tell us about your approach to real-world applications.

Kush: In approaching real-world applications, it is critical to truly understand the problem by having deep and continual interactions with subject matter experts.  Superficial understanding leads to superficial solutions, but in-depth understanding gives a chance at truly making a dent.

Saška: That is why we chose to partner with NGOs in our program. They are the folks who are at the forefront, working to alleviate hunger, poverty, inequalities. To scope the six projects for our first social good season in 2016, we interviewed over 150 NGOs. We learned from every single one.

5) When will we see the next generation of tools for consumers to use AI to lead healthier lives?

Kush: One of the big promises of AI in health and wellness is the facilitation of a high level of personalization and precision for patients. I think we’re starting to get there.

6) Tell us something surprising about yourself that keeps you motivated beyond data!

Saška: When I was six years old, my dad gave me his trusted Kiev film camera and an expensive light meter. Forty years fast forward, photography is still a big part of my life; whenever I press the shutter, it feels as if I preserved a precious little piece of this fast-moving world.

Kush: In 1908, my great-great-grandfather started up what we would today call a high-tech crowdfunded social enterprise: the Paisa Fund Glass Works. If he could walk the fine line of both “doing good” and “doing well” then, I can surely give it a try now.

The LAUNCH Food network is a global, cross-sector coalition of stakeholders committed to creating positive change in the food system. Are you interested in joining? Email