Artificial intelligence is everywhere. But behind the promise of autonomy lies a more complex reality: AI relies on humans. My recent research explores this often-overlooked fact, showing how people remain central to how AI functions as active contributors, far exceeding their portrayal as passive users.
In a recent MIS Quarterly (open access) publication, I led a study of algorithmic autonomy in ride-hailing platforms like Uber. These systems often appear seamless, powered by algorithms that dispatch drivers, set prices, and optimize routes. But beneath the surface, we found more nuance. It is not a single algorithm making decisions, but a network of algorithms that delegate activities to humans (drivers, passengers, or platform workers) in distributed and sometimes nearly invisible ways.
We call this phenomenon distributed delegation. Rather than replacing human acting, platforms redistribute it, embedding it in interface design, ratings, and optimization rules. The result is a hybrid form of acting where algorithms and humans are co-dependent: machines rely on human data, choices, and labor to function effectively, while shaping how those humans act and what they can do. These findings challenge the dominant narrative of AI autonomy. We reveal that algorithms don’t automate away human involvement; rather, they restructure it, often in subtle ways.
Building on these insights, I’m currently conducting research that explores humans and algorithms acting together in multi-faceted interactions. Rather than asking who holds control, I focus on uncovering how the capacity to act emerges through interaction between actors. I study a case of an artificial pancreas system that combines continuous glucose monitoring, insulin pumps, and user-configured algorithms to automate insulin delivery. This case offers a revealing insight into the idea that acting in AI-enabled systems is not something done by actors separately, but it emerges from ongoing interactions between humans, bodies, algorithms, data, devices, and the environment.
Across industries, AI promises speed, scale, and precision. Yet, my research shows that its success still depends on human roles. Whether in platform work or healthcare settings, humans interpret, correct, and adapt algorithmic systems in ways that shape outcomes.
Designing Human-AI Work
For business leaders, this means we must move beyond the myth of full automation. The challenge isn’t simply how to deploy AI, but how to design and manage human-AI interactions, supporting the people whose actions and decisions make intelligent systems work. As algorithms take on more decision-making functions, workers are often tasked with invisible, fragmented, or reactive forms of labor: interpreting machine outputs, fixing edge cases, or providing the flexibility that automated systems can’t achieve. This labor is essential but often goes unrecognized and unsupported. My research suggests that valuing and designing for this hidden work is a key leadership challenge. It requires rethinking how we train, support, and reward workers who now collaborate with machines.
For those just entering the workforce, my research carries a hopeful but strategic message: AI still needs you. The most effective systems aren’t fully autonomous but rather they rely on human insight, adaptability, and judgment. As intelligent technologies become more integrated into business processes, your value will lie in your ability to work with and around machines—to interpret outputs, ask hard questions, and make context-sensitive decisions. Don’t just aim to keep up with AI. Instead, develop the human skills that machines can’t replicate: critical thinking, relational sensitivity, and awareness at a systems’ level. Those are the capabilities future workplaces will increasingly demand.
As AI systems grow more pervasive, the real challenge isn’t replacing human intelligence—it is understanding how to build with it. My research in this area continues to highlight how digital systems and human contributions are deeply entangled. For alumni and future leaders, the opportunity lies in shaping technologies and organizations that do not just automate but emphasize human strengths. A more intelligent future isn’t just about machines that think better, but about systems that help us all think, work, and live more wisely.
About Marta Stelmaszak Rosa, PhD
Marta Stelmaszak Rosa is an assistant professor in Operations & Information Management at the Isenberg School of Management at UMass Amherst. Her research centers on digital data and their responsible, sustainable, and socially just management in organizations. She teaches courses on business intelligence and data analytics. Stelmaszak Rosa’s work on designing human–AI interactions is shaping Isenberg’s evolving curriculum, from newly launched undergraduate and graduate AI certificates to expanded elective options across degree programs.
© Copyright 2025 Poets & Quants. All rights reserved. This article may not be republished, rewritten or otherwise distributed without written permission. To reprint or license this article or any content from Poets & Quants, please submit your request HERE.
