From NASA to the Enterprise: LISH Expands Programs in AI, Crowdsourcing, and More

April 2, 2021
LISH logo

by IQSS Staff

The Laboratory for Innovation Science at Harvard (LISH), one IQSS’s most dynamic scientific programs, has experienced unprecedented growth and accomplishment over the past two years. While physically located at the Harvard Business School, LISH has enjoyed a partnership with IQSS dating back to 2009, which has enabled the lab to leverage the expertise of IQSS’s HR, finance, and sponsored research administration teams to bring new collaborators into the organization from all over the world and secure funding from a variety of external sources, including NASA, “Eric and Wendy Schmidt Foundation, and the Linux Foundation. The relationship of more than a decade culminated in a recent 18-month period during which LISH published 28 articles, papers, and cases, developed three courses, hosted 19 new visitors, and welcomed HBS Professor Marco Iansiti as a faculty co-director of the lab. To say LISH has been busy would be an understatement. 

The Lab and COVID

Jin Paik leaning against a wood pillar in front of a stone building
LISH director Jin Paik

Since the beginning of the COVID-19 pandemic, LISH has made strides in research by better understanding behaviors from scientists—both in education by re-launching an open online course of technology entrepreneurship, and in practice by designing virtual workshops for the use of artificial intelligence (AI) in the Enterprise. 

We sat down with Jin Paik, director and senior researcher at LISH, to find out more about programming and the initiatives that LISH is spearheading in six areas of application: Crowdsourcing & Open Innovation; Data Science & AI Development; Science of Science; Technology Commercialization; Business of Sports, and Networks.

“Our mission is to spur the development of a science of innovation through a systematic program of solving real-world innovation challenges while simultaneously conducting rigorous scientific research and analysis. We see the world as our lab. We conduct rigorous A/B Testing with organizations that are experimenting with innovation in the field. With that, we look to have an impact on practice.”  

Nearly 70,000 students have taken their massive open online course on Technology Commercialization on HarvardX, with the vast majority enrolling after March 2020. Students in this self-paced course discover pathways to match technologies with potential consumer needs. Entrepreneurs can think critically about the gaps in the market and how to bring lab technologies into real-world application. 

Paik reflected on the popularity of the course, saying, “It’s truly remarkable to see the level of engagement out there. We know the contents of the curriculum were strong, but the response to the global community has signaled that people are eager to transform the development and use of advanced technologies like IoT (Internet of Things), 3D printing, AI, and smart sensors.”

The pandemic has brought on different challenges for the academic community. LISH’s Science of Science track aims to understand how labs operate, what makes them productive, and the drivers, behaviors, and motivations behind innovative work. LISH recently ran a survey with 4,535 faculty or PI respondents on the impact of the COVID-19 pandemic. "Unequal effects of the COVID-19 pandemic on scientists,” published in July 2020 in Nature Human Behaviour, found that female scientists with children at home were experiencing the largest decline in productivity (Myers et al. 2020). 

Jin Paik, wearing a surgical mask, standing in front of the LISH logo on a wall in an office
Jin Paik tours LISH's new office on Harvard's Allston campus

Collaborative Roots

LISH was created in 2009—originally called the NASA Tournament Lab—by Professor Karim Lakhani “as a joint initiative between NASA, Harvard Business School, and IQSS to work on designing competitions or tournaments to create the best computer codes and data analytics solutions for NASA and other governmental agencies.” Since that time, LISH has helped organizations run open innovation programs at scale. 

Paik said, “Between our work with NASA, the Harvard Medical School, the Broad Institute, and other firms, we have run over 750 discrete innovation contests over the past ten years. We’ve had a large impact on how crowd-generated algorithms often perform better than internal benchmarks. In the most recent collaboration with the Broad Institute, we turned to the crowd to develop an algorithm to predict a drug compound’s mechanism of action given its cellular signature.” Attracting 4,000 teams resulting in over 72,000 submissions, preliminary analysis of the results show that 20% of the teams improved the Broad’s prior efforts by at least 8–10%, with a max improvement of 15–20%. These new algorithms will help scientists advance the drug discovery process. 

Under the Data Science and AI track, Paik explained how LISH is exploring multiple facets of AI development, “We want to understand how this stuff gets deployed at large companies. Is digital transformation the baseline? What is the next thing? Automating? Who is making predictions off the automation?” The approach LISH has taken is two-pronged, first focusing on knowledge-sharing with a monthly practitioner series, Artificial Intelligence in the Enterprise, where industry experts share experiences and strategy on the implementation of AI in large organizations with the practitioner community. 

Second, LISH is conducting empirical research on the adoption of machine learning and AI in the firm. Through field research and experiments, LISH researchers are addressing questions such as human + machine performance, differences in judgment, transformation in firm operations, and explainability and fairness of models. Paik expanded on one of LISH’s current partners, Commonwealth Bank: “They built a customer service engine that they have used over the last five years. Online, they take all personal data from calls to retail, chatbox, online—and when customers go into the branch, it’s all streamlined because the back engine is AI, thereby predicting what service or product will best meet your needs at that given moment. In determining which messages will have the highest success rate, technical and frontline staff work together to run experiments to determine which products to offer customers and measure whether or not it makes sense or works.” 

Sponsors and Partners

LISH has a track record of working with outside organizations, as it has received funding from NIH, HBS, Alfred P. Sloan Foundation, NSF, NASA, and the MacArthur Foundation, among others, and has partnered with various organizations and companies, including NASA, the US Department of Defense, Topcoder, Kaggle, Broad Institute, Dana-Farber, and Deloitte. In 2020, LISH received a $750,000 gift from Woodside Energy, the largest producer of natural gas in Australia. LISH is researching how the application of open innovation tools, through contests for technology solutions, will assist Woodside in developing ideas and technology for blue and green hydrogen and robotics. Academic and industry leaders learn how to get new energy faster.

For more information about the Laboratory for Innovation Science at Harvard, you can visit