Rethinking Organizations for the AI Era

The question isn’t whether AI will transform organizations, but whether yours will be ready.

Welcome to FullStack HR, and an extra welcome to the 98 people who have signed up since last week.

If you haven’t yet subscribed, join the 8600+ smart, curious HR folks by subscribing here:

---

Happy Wednesday,

We're just a few weeks away from HR Tech Amsterdam, and I still have six of the Johannes Special Tickets left. If you want one, just email me—first come, first served!

Before we dive in, I have a small favor to ask.

This newsletter reaches between 10,000–20,000 readers, but only a handful take a second to click the heart.

If you find value in what I share, consider tapping it today—it’s a small gesture that helps more than you might think.

Now, onto today’s piece.

This one is a bit longer than usual because it’s an article I’ve been working on for quite some time. ChatGPT Deep Research has been incredibly useful in challenging my thoughts, providing insights, and expanding my perspective on the topic. It’s marvellous to have a research at hand.

This isn’t a “do this” kind of article; rather, it’s meant to spark new ideas and, hopefully, some discussion.

Let’s get into it.

---

The last two years, I – like many others – have focused on individual AI skills development. While this remains crucial, especially in the Nordics where we lag behind, I believe we need to expand our vision to what AI means for organizational structure itself.

My firm belief is that the companies that crack what an AI-driven organization will look like first are the ones that will succeed in this new era.

Perhaps we've spent so much time worrying about whether individuals can adapt to AI (I'm partially guilty for this focus), but what about the actual structure of organizations? If your company is built like a rigid hierarchy with slow decision-making and siloed teams, no amount of AI will save you.

AI thrives in flexible, agile (!), networked organizations.
From my point of view, that's where the real potential lies.

The Network Organization – Built for Speed and Adaptation

Traditional organizations are slow. Information crawls through layers of management, roles are fixed, and innovation gets stuck in bureaucratic approval loops. That's a problem because AI doesn't wait.

It processes, predicts, and allocates in real time. Or soon will.

We tend to project human-like qualities onto AI, assuming it will function as a tool that simply helps us work more efficiently.

But that's the wrong way to think about it.

AI won't just match human capabilities—it will surpass them. And that's where most people get it wrong. They assume AI will fit neatly into existing structures, but AI doesn't work within human limitations. It doesn't wait for meetings, it doesn't hesitate, and it certainly doesn't follow our traditional corporate logic.

This is where organizations will trip up.

AI isn't just a faster worker or a better assistant—it's something fundamentally different. It doesn't see departments, hierarchies, or even processes in the way we do.

It identifies patterns, optimizes without sentiment, and redefines efficiency in ways we haven't even thought of yet. If we continue treating AI as just another tool, rather than the foundational shift it represents, we won't just lag behind—we'll be irrelevant before we even know what happened.

So, the real question isn't how we use AI, but rather how we build for AI. And that starts with how we organize, how we distribute decision-making, and how we allow AI to operate not just as a tool but as an integrated force in shaping the future of work.

The Evolution of Network Organizations

When I studied HR, Walter Powell's ideas about network organizations challenged conventional wisdom. His 1990 paper, "Neither Market Nor Hierarchy: Network Forms of Organization," presented a vision that seemed almost too fluid to be practical.

In industries where knowledge is the key asset, Powell argued that firms don't operate solely through hierarchical control or market-based transactions. Instead, they rely on network structures—collaborative, trust-based relationships that foster innovation and adaptability.

I saw elements of this at Spotify—agile ways of working, squads, tribes, and all that. While the agile movement sometimes felt like it was running at full speed without questioning its fit for every situation, Powell's insights now seem prescient for the AI era. Unlike traditional hierarchies that depend on rigid workflows and bureaucratic oversight, network organizations allow tasks and relationships to shape the structure dynamically.

AI: Beyond Tool to System

AI's biggest strength isn't just automating routine tasks – it's optimizing collaboration. Imagine an organization where AI doesn't just handle logistics but actually makes real-time decisions about how work should be done and by whom. The real power of AI isn't just in efficiency; it's in its ability to dynamically match the right resource to the right task, whether that resource is a human or a machine.

AI doesn't just work within the system—it is the system. It has the potential to not only outperform human capabilities in certain areas but also to act as the orchestrator, distributing work to the right "player" at the right time. This isn't theoretical—this is happening right now.

Look at how AI is already transforming biotech. Cradle's generative AI platform is accelerating protein engineering by identifying the best experimental protocols and automating routine testing, allowing scientists to focus only on the most complex cases. Early adopters report reducing development cycles by up to 50% while improving success rates.

The same shift is happening in customer experience. Odigo's AI orchestrator routes millions of customer interactions, seamlessly balancing tasks between AI chatbots and human agents. Their clients have seen average handling times drop by 30% while maintaining or improving customer satisfaction scores.

These are not "just" efficiency gains; they are shifts in how work is allocated and executed.

The Future of Work: Opportunities and Challenges

Network organizations also gain from AI's ability to scale work up or down effortlessly. They don't need to carry a massive full-time payroll when AI allows them to plug in external resources on demand.

But this shift comes with a challenge—what happens to traditional employment structures when organizations can swap workers for AI-driven solutions on an as-needed basis?

The traditional job market, with its emphasis on long-term employment and career progression, will be fundamentally disrupted. Job stability may decrease, replaced by a more fluid, task-based economy where adaptability is key. Companies that embrace this will thrive, but for individuals accustomed to stable, structured employment, this transition could and will probably be daunting.

The Social Contract at a Crossroads

Let's pause here for a moment. We're discussing these organizational changes as if they're inevitable – and technologically, they might be. But we're facing a fundamental choice about the kind of society we want to build. The stable employment model wasn't just about organizational efficiency; it was a social contract that provided people with dignity, purpose, and the ability to plan their lives.

When we talk about shifting to a fluid, task-based economy, we're not just changing business models – we're potentially dismantling a social framework that has supported families and communities for generations.

Yes, the traditional employment model had its flaws, but it offered predictability, healthcare benefits, retirement planning, and a sense of belonging.

Consider the human cost:

- How do people plan for families or mortgages in a completely fluid work environment?

- What happens to mental health when job security becomes a thing of the past?

- How do we maintain social cohesion when the workplace no longer provides a stable community?

- What about those who aren't among the "elite talent" – do we simply accept their displacement?

The Nordic model has long balanced market efficiency with social welfare. Perhaps instead of just asking how to build AI-driven organizations, we should be asking: How do we harness AI's potential while preserving or reinventing the social benefits that traditional employment provided? Could we create new forms of security and belonging in a fluid economy?

This isn't just about making organizations more efficient – it's about the kind of society we want to live in. The companies that truly succeed in this transition might be those that find ways to combine AI-driven flexibility with new forms of stability and security for their workforce.

The Real Question: Who (or What) Should Do What?

So how should organizations structure themselves to make the best use of AI while still leveraging human capabilities where they matter most? As said, I think the companies that figure this out will dominate whatever industry they happen to operate. Long-term, those that don't will struggle, weighed down by hierarchies, outdated job structures and slow decision-making.

At the same time, there are risks. AI-driven networks could concentrate power in the hands of a few dominant players, forcing smaller firms to rely on external AI infrastructure owned by giants like Google, Amazon, and OpenAI. Then as mentioned above, the death of job security which is a real concern—many networked organizations will prioritize cost savings over long-term employment, contracting only the most elite talent while giving everything else to AI.

If AI can generate code, mid-level developers may find themselves squeezed out, just as AI-powered legal analysis might reduce the need for junior lawyers.

AI bias and ethical dilemmas also come into play. If AI decides who gets work and who doesn't, systemic bias could become invisible and unchallenged, leading to inequalities that are hard to fix.

Building for the AI Future

Before you jump into AI adoption, you need to take a serious look at your organization's structure. Is it designed to flex and adapt, or is it built on outdated hierarchies that slow everything down? If your AI strategy is just about plugging automation into an old model, you're missing the point. AI demands a structure that can keep up with it.

Trying to integrate AI into a traditional corporate structure is like swapping a steam locomotive's boiler for a jet engine and expecting it to fly. The whole structure needs to change. And that's where network organizations shine.

The businesses that get this right – the ones that rethink their structure, embrace flexibility, and learn how to deploy AI strategically – will outcompete everyone else.

I think AI-powered network organizations will win.

The only question is whether your company will be one of them.