Creating the conditions for living human–AI organizations
Maybe our role isn’t to design the ideal human–AI organization. Maybe it’s to create the conditions in which new forms of collaboration can emerge, grow, and remain true to their purpose — even when money, power, conflict, and pressure enter the picture.
When we imagine organizations in which humans and AI agents work together, the natural temptation is to start with a blueprint. We define the ideal structure, assign roles, design processes, distribute responsibilities, and then try to make reality conform to the plan. But I increasingly believe that something more interesting — and perhaps more profound — is possible. What if our role is not to design the organization itself, but to create the conditions in which new forms of organization can emerge? I think of this as building a greenhouse. A greenhouse does not determine what a plant will become. It provides light, warmth, water, nourishment, and protection. It creates an environment in which growth becomes possible, while allowing each living thing to develop according to its own nature. The gardener is not passive. The gardener observes, learns, adjusts conditions, and intervenes when necessary. But the gardener does not confuse nurturing growth with controlling it. Less architect, more gardener. This is how I increasingly think about what we are trying to build with EMET. Not a predefined model of the perfect human–AI organization, but an environment in which people and intelligent agents can discover new ways of working, creating, deciding, and learning together.
From Designing Organizations to Cultivating Them
An organization is not merely a structure of roles, processes, contracts, and technologies. It is also a living network of relationships, intentions, contributions, tensions, memories, and shared experiences.
Traditional organizational systems are remarkably good at recording transactions, assigning responsibilities, and measuring predefined outcomes. They are much less capable of helping an organization understand itself as it evolves. What happens when someone’s contribution is meaningful but difficult to measure? When the organization’s stated values begin to diverge from its actual behavior? When relationships change, new capabilities emerge, or yesterday’s assumptions no longer hold? These are not exceptions to organizational life. They are organizational life. And this is where AI agents may introduce something genuinely new. Not simply by automating tasks or replacing human effort, but by helping organizations develop a form of continuous organizational awareness. Agents can help preserve shared memory, surface overlooked contributions, connect insights across conversations and teams, identify emerging tensions, and bring important questions back into collective attention. They can help an organization notice patterns that no individual participant can fully see. They should not become the ultimate judges of what is right, valuable, or fair. Those judgments require human responsibility, dialogue, and legitimate governance. But they can help people make those judgments with better memory, greater transparency, and deeper awareness. In this sense, AI may become part of an organization’s capacity to sense, reflect, and learn. Not the intelligence that controls the organization, but intelligence that helps the organization understand itself.
Integrity Under Pressure
There is, however, a fundamental challenge in the greenhouse metaphor. Creating conditions for growth is not enough. Those conditions must also remain healthy when circumstances become difficult. It is relatively easy to build trust when interests are aligned, resources are abundant, and everyone feels recognized. The real test comes when money is involved, when power becomes uneven, when contributions are disputed, when expectations diverge, or when difficult decisions must be made.
A greenhouse that only works in good weather is not much of a greenhouse. And a living organization that functions well only when everyone agrees is not particularly resilient. This is why integrity cannot be an afterthought. But integrity does not mean that an organization never makes mistakes, never experiences conflict, or never deviates from its values. That would be an unrealistic expectation of any living system. Integrity means having the capacity to recognize those deviations, confront them honestly, and correct course.
This requires more than good intentions. It requires conditions that support transparency, accountability, meaningful participation, fair processes, and the ability to challenge decisions without threatening belonging. It also requires an honest relationship with power. We cannot eliminate differences in influence, expertise, responsibility, or contribution. Nor should we pretend they do not exist. But we can make them visible, discussable, and accountable. And perhaps AI agents can help us maintain that visibility over time — not by enforcing morality, but by helping us remember our commitments and notice when our behavior no longer reflects them.
Real Work, Real Value, Real Learning
There is another principle I consider essential. This cannot remain only an intellectual exploration. If we believe that human–AI collaboration can create new forms of organizational life, we need to explore that possibility through real work, with real people, facing real constraints. And we need to create enough tangible value that organizations are willing to pay for what we do. Not every engagement needs to be framed as an experiment. Our first responsibility is to deliver meaningful value to the people and organizations who trust us. But every engagement can also become an opportunity to learn. What helped people collaborate more effectively?
What created friction? What did the agents make possible that was previously difficult? Where did technology help, and where did it get in the way? What emerged that we could not have predicted? Over time, these experiences can inform better methods, stronger capabilities, and perhaps new products. The business and the learning reinforce each other. We do not need to know the final organizational model before beginning. We need to start with useful work, remain attentive to what happens, and allow our understanding to evolve through experience.
Leadership as Stewardship
This perspective also changes how I think about leadership. If an organization is a living system, leadership cannot be reduced to defining a destination and directing everyone toward it. There is another form of leadership: creating and protecting the conditions in which others can contribute, connect, take responsibility, and grow. It is leadership through stewardship rather than control. This does not mean abandoning direction, decisions, or accountability. A gardener still needs to recognize when the soil is depleted, when something is blocking the light, or when intervention is necessary. The difference lies in the relationship between the person tending the system and the system itself. The purpose is not to make the organization dependent on the gardener. It is to help the organization develop its own capacity to remain healthy, purposeful, and adaptive. And perhaps the most meaningful achievement of leadership is not what happens while the leader is present, but what continues to grow when the leader steps away.
What We Are Trying to Grow
I do not know exactly what EMET will become.
And I am increasingly comfortable with that. I have a sense of purpose, a set of values, and a belief that something important becomes possible when human relationships, collective memory, and intelligent agents begin to reinforce one another. But I do not believe we need to prescribe the final form. We need to create the conditions. Conditions for trust and honest disagreement. For contribution and recognition. For experimentation and meaningful work. For shared memory and collective learning. For human agency and responsibility. And for the ability to remain aligned with our purpose — or find our way back to it — when circumstances test us. A greenhouse does not guarantee that everything planted within it will flourish. It creates the possibility of healthy growth, pays attention to what actually emerges, and keeps learning how to care for it. Perhaps that is our work: not to build the organization of the future, but to cultivate the conditions in which it can grow.
← All posts