Researchers Worldwide Compete to Shape the Future of AI in Organizations

As AI tools become commonplace in the workplace, the way teams form and collaborate will evolve. Organizational models must adapt to support innovative ways of working, but it’s hard to predict what shape they’ll take.

To explore this emerging field, Stanford HAI collaborated with Google DeepMind to launch the AI for Organizations Grand Challenge, an invitation for scholars worldwide to submit their best ideas for studying the future of collaboration within organizations. 

“Getting to work with a frontier lab that wants to understand how AI is reshaping the workplace is a rare opportunity for scholars in this field,” says Melissa Valentine, HAI senior fellow and associate professor in the management science department at Stanford. “This competition marks the beginning of a broad, public conversation about how organizations are changing.”

More than 200 teams of faculty and PhD students from 156 universities responded to the challenge, submitting proposals that spanned three categories: using AI as a tool for improving alignment in organizations, understanding the human impact of deploying AI in organizations, and simulating the behavior of teams with synthetic organizations.

A hybrid panel of judges from six leading universities and Google DeepMind leaders evaluated the proposals for novelty, impact, and feasibility in a double-blind review process. After the initial evaluation, 13 teams were invited to pitch their ideas to the panel.

First Place: Learning the Grammar of Coordination

The winning submission from Yankai Wang, a PhD student in the Stanford Graduate School of Business (GSB), and Amir Goldberg, professor of organizational behavior at GSB, proposed to study the “grammar” of coordination.

“Coordination unfolds through a series of human interactions. People send emails, hold meetings, edit documents – but we don’t know what makes a sequence of actions effective in one situation versus another,” Wang says. “We want to develop a framework to help leaders understand the dynamics of coordination and make decisions grounded in organizational science instead of having to trust someone’s instinct.” 

To this end, Wang and Goldberg will use the modern transformer architecture for machine learning, which excels at finding patterns in sequential data, to build a “large coordination model” that learns how successful teams coordinate their work and predicts which sequence of actions will work best for a given scenario in the future. 

As part of the $100,000 grand challenge prize, the winning team will be able to implement its study in Google DeepMind’s own offices. The company will collaborate with the researchers by providing compute time, engineering resources, and mentorship.

“With a novel approach to understanding coordination, we can unlock new theoretical and practical opportunities in different types of organizations,” says Martin Gonzalez, head of organizational AI research for Google DeepMind.

Finalists Propose Using AI to Enhance Human Collaboration

Four additional finalists were recognized at the AI for Organizations Conference on May 12:

• Lean Curation: Managing Organizational Innovation in the Age of Cognitive Abundance – A team of scholars representing Emory University, Cornell University, and Carnegie Mellon University envisioned applying the concept of lean manufacturing to organizations to help them decide which of the many ideas generated by AI is worth pursuing with time and resources. 

• Co‐AI: Using AI to Strengthen Collective Intelligence in Organizations – A team from the Tepper School of Business at Carnegie Mellon University proposed a new way of using AI to measure collective intelligence in teams. 

• From Invisible to Accessible: AI-Fueled Collaboration and the Discovery of Organizational Expertise – A group from the Haas School of Business at the University of California Berkeley and INSEAD imagined using AI-fueled recommendations to help teams surface internal expertise that’s trapped in siloes across the organization.

• TeamLens: Multimodal Large Language Models for Surfacing Team Collaboration in Real Time – Researchers from the Kellogg School of Management and Northwestern Institute for Complex Systems at Northwestern University proposed to link team science theory with multimodal LLMs that can analyze team behavior at scale to improve collaboration through a better understanding of the factors that determine success.

The AI for Organizations Grand Challenge is part of a multi-pronged effort to make sure workplace transformation happens in a human-centered way. In a related development, Stanford HAI announced the new AI and Organizations Lab, led by Valentine, with funding from Google DeepMind.

“The field of organizational science is moving faster than most people realize,” says Simon Bouton, Chief Experience Officer, Google DeepMind. “We look forward to continuing this collaboration to shape the future of AI in organizations.”