6

Final Discussion and Synthesis

Sharon Kirkpatrick began with a review of the workshop’s goals, which were to discuss promising opportunities and directions and best-known practices in the application of advanced computation, big data analytics, and high-performance computing to food and nutrition research; consider likely pitfalls, including those associated with privacy, bias, and trust, and safeguards to avoid them; reflect on the appropriate use of evidence generated by these methods; and consider needed investments in capacity development. She summarized some of the workshop’s main themes:

• Applying artificial intelligence (AI), machine learning (ML), and deep learning (DL) to food and nutrition research is a series of interrelated moonshots, with both optimism and pessimism that these moonshots will succeed
• Making existing data more available, improving the representation of ongoing data collection, and addressing issues of data sovereignty
• Assembling diverse teams that collaborate and communicate effectively and develop a shared vocabulary and culture
• Engaging communities and other stakeholders early in study design
• Developing and using AI/ML/DL tools in a manner that supports equity, fairness, and justice for all end users and reflects the broader context of social and structural determinants of health and nutrition
• AI is useful, but it is not magic and requires human expertise to be most useful

Christopher Mejía-Argueta agreed with Kirkpatrick’s summary and emphasized the importance of ensuring that technology, data, and model-driven strategies include input from stakeholder communities and will benefit the smallholder farmers, retailers, and family-owned businesses, which will have a role in including nutrition in the food ecosystem. Sai Krupa Das noted that the research community has developed useful AI-enabled tools and is smart enough to recognize that data are messy and have gaps and to work together to fix this. With those tools, the field can develop the infrastructure that will enable moving toward the end goal of improving human health while benefiting the planet. She recognized the possible legal and ethical pitfalls that Janie Hipp discussed and the importance of ensuring that data and models are not exploited in ways that harm agricultural producers by making a concerted effort regarding data surveillance, data monitoring, stewardship, and equitable access.

Angela Odoms-Young wondered how AI could help shape food assistance and production policies so they reflect the differential responses that various subpopulations have to food intake and nutritional advice. By focusing on subpopulations rather than individuals, such policies may transform the structures that contribute to inequity.

Rodolphe Barrangou reiterated that foundational AI tools exist and, although they can be adapted and improved, they have established a sound platform and basis for use. The main limitations in his view concern the data. Yes, data are abundant, but they are not always accessible, formatted properly, of the highest quality, curated, or regulated correctly. He also expressed concern about how those working in the field in different disciplines can better collaborate and develop a common language and culture and said that the field needs to involve nutritionists, clinicians, data analysts, food scientists, farmers and ranchers, public relations experts, social scientists, behavioral scientists, economists, supply chain experts, and marketing and communications experts. He called for incentivizing people from these disciplines to join the teams needed to harness nutrition and food to preventively manage health as opposed to curing disease. “I think we need to strategize about the who and the how we are going to work together to be more efficient,” said Barrangou.

Kirkpatrick noted the general lack of institutional support for team-based science and wondered how to address this problem. Mejía-Argueta acknowledged that this is a difficult situation and suggested ways to bolster team performance:

• Identifying the key performance indicators for each member and similarities across the team that can be leveraged to benefit everyone regarding institutional recognition, perhaps by enabling different members to serve as the primary author on publications reporting the findings.

• Using grant funds to collaborate with colleagues from countries outside of the United States that face similar problems and challenges.
• Inviting people from industry or nonprofit organizations to provide feedback on programmatic activities to expose team members to the need to ensure that the team’s work needs to be useful.
• Gaining the trust of others to create alliances that are sustainable over the long term and can create their own key performance indicators.
• Developing a culture that creates synergies and ensures the team is robust and able to address the same problem from different perspectives.

Das noted that everyone may not be equipped or would need to start assembling large teams but could be ready to contribute to efforts that would benefit from their domain expertise and that this will require keeping up with and understanding the landscape and comprehensive nature of ongoing research and providing expertise within that context. Odoms-Young noted the movement at some institutions to change the key performance indicators used for promotion and tenure decisions to reflect the value of team-based research or community-based participatory research, for example. The key is having the right metrics of success in place, which perhaps professional organizations could develop to reflect new ways of engaging in research.

Barrangou agreed about changes to promotion and tenure metrics and called for the community to build new infrastructures to support this growing field. Odoms-Young added that policy makers have also begun to change how they decide on where grant funds go, but the field needs to do more to engage them and the public to show the value that team science creates. Das said that the tide is changing in a way that empowers trainees and early career investigators to push for new infrastructures and new norms regarding promotion and tenure.

Aaron Smith said that deep disciplinary expertise is needed for the field to realize its potential. Therefore, Smith continued, it is important to ensure that training does not dilute that by only producing people who are good at everything; generalists help disciplinary experts talk to one another.

This page intentionally left blank.