This article was first published in our Cocooners N° 17 and is part of the How we work series.
If you would like to receive updates on this and other topics related to the evolving world of work, subscribe to our newsletter here: https://cocoon-pro.com/#subscribe
To find out more, or get your organisation involved in this approach, contact us here: https://cocoon-pro.com/get-in-touch/
Superlatives often dominate the current conversation around artificial intelligence. There are reasons for that: studies commissioned by Google, for example, estimate that generative AI could add €1.2–1.4 trillion to the EU’s GDP—an 8 % boost in a decade, and more than six in ten jobs are expected to collaborate with AI. Yet a more fundamental question underlies these numbers: what happens when a work team becomes augmented? Beyond chatbots that answer prompts, AI agents are emerging—autonomous systems that communicate, decide, and learn. They are designed to take action, adapt to new situations, and solve problems without constant human supervision, bringing social capability, autonomy, reactivity, and proactivity. Unlike generative models, agents execute tasks and evolve over time. Their arrival poses both hopes and challenges for the entire work ecosystem. What happens as they now become members of mixed human + AI agent teams?
After fifteen years of studying and improving team performance, we have now launched a specific action‑research initiative to explore this very question. We are partnering with research institutions and a variety of organisations to learn in practice. We aim to generate shared insight on how augmented teams function and how to ensure they benefit both people and organisations. We start from a socio‑technical perspective: team performance arises from the interaction of social processes, technology, and structure.
How do we divide work between humans and agents while preserving engagement and fairness? How must communication and coordination evolve? How do we build trust and manage the psychological impact on autonomy and security? Who is responsible when a machine influences a decision? How do feedback loops allow both humans and AI to learn together?
These questions are grounded in real cases. In each case, technology amplifies human insight instead of replacing it. These are uncharted waters, and we are committed to learning openly and sharing what we learn. Action-research is an iterative, collaborative process. It produces knowledge that is relevant to real contexts. Our initiative builds on this tradition, adding the strength of our partners across academia and industry. By inviting heads of HR, COOs, entrepreneurs, and team leads to join us, we hope to co‑create guidance that the whole ecosystem can use.
If you are curious, concerned, or simply eager to learn, we invite you to connect. Together we can raise the signal above the noise and build a work ecosystem where human and artificial intelligence complement each other, helping teams thrive in ways we are only beginning to imagine.
This article was first published in our Cocooners N° 17 and is part of the How we work series.
If you would like to receive updates on this and other topics related to the evolving world of work, subscribe to our newsletter here: https://cocoon-pro.com/#subscribe
To find out more, or get your organisation involved in this approach, contact us here: https://cocoon-pro.com/get-in-touch/