The Architecture of Power: From Steam to Silicon

By Nathan Waterhouse · 2025-01-15

Throughout history, the physical and organisational structure of work has been intimately tied to the prevailing source of power—both literal and metaphorical. AI empowers employees in ways no other technology has in history. How will it change organisational dynamics – and what structures might best equip employees to take advantage of its potential to create positive impact for them, shareholders, and society as a whole?

First, let's take a look back in time to look at how tech has changed org structures.

The steam engine significantly boosted productivity, allowing for mass production and the rise of factories. This shift away from small-scale, artisanal production enabled larger workforces and necessitated more complex management structures. Steam powered factories had centralised layouts, as power could not easily be distributed over long distances. Workers clustered around these mechanical hearts of industry, their movements and routines dictated by the rhythmic pulse of steam-powered machinery.

When the power source switched to electricity, factory design didn't immediately change. Instead, the steam engine was just switched to electric. Factory management didn't immediately see the potential for completely new ways to organise labour. But eventually, with the birth of the assembly line, this led to the first great decentralisation. No longer tethered to a single power source, factories could fragment into specialised departments, each with its own electrical infrastructure. This physical transformation caused a deeper organisational shift: the rise of functional specialisation and departmental silos that would define 20th-century management theory.

The Knowledge Revolution and Its Paradoxes

The advent of the internet ushered in our second great decentralisation—the democratisation of knowledge. A student in rural Kenya could now access the same information as their counterpart at Oxford or Harvard. The internet enabled real-time access to shared information, encouraging open physical spaces to symbolise and support open information flow. As knowledge work grew more "open," it also led to information overload and new forms of micromanagement via digital tools.

AI and the Third Decentralisation

We're now witnessing a third wave of decentralisation, but this time it's not just knowledge being democratised—it's capability itself.

Richard and Daniel Susskinds' observations about the medical and legal professions prove particularly prescient here. When AI systems can match or exceed human experts in diagnostic accuracy or contract analysis, the traditional monopoly of expertise begins to crumble.

Consider these emerging patterns:

Reimagining Organisational Structure

The traditional organisational pyramid, borrowed from military hierarchies, was designed for a world where information scarcity necessitated careful control and distribution. But what happens when AI makes expertise increasingly accessible?

Of course, we don't yet fully understand AI's impact on organisations–it's too early. As we saw with the electrification of factories, with every dramatic change in technology or power source, it takes time for organisations to change shape, and those holding the threads of power often resist changes that will loosen their traditional command and control.

AI's deployment doesn't only hinge on technological readiness but on trust, transparency, and change management—see my previous blog post on the importance of culture. Given we don't have good examples of AI-native companies yet, perhaps we can look for examples of an increasing flattening of organisations with greater autonomy.

Emerging Models: Decentralised Organisations in Action

Teal Organisations

Buurtzorg, a Dutch healthcare organisation, demonstrates the power of decentralised authority. With over 14,000 employees and 900 teams, they operate without traditional managerial layers. Their digital platform, Buurtzorgweb, streamlines operations and enhances team autonomy, contributing to a 40% reduction in overhead costs and the highest patient satisfaction scores in the Netherlands.

As a shining example of a Teal organisation, Buurtzorg showcases how self-managed teams can drive both efficiency and quality. Other notable examples include Patagonia, FAVI, and Sun Hydraulics.

Chinese Digital-Enhanced Direct Autonomy

There's been a growth of companies in China which offers a particularly fascinating insight into radical organisational transformations.

Handu Group's transformation of fashion retail through its HStyle division showcases the power of radical decentralisation. Their micro-teams start typically as only 3 people and combine creative and commercial roles—designers working directly alongside inventory specialists and web experts—creating autonomous units that can respond to market trends with extraordinary speed.

What makes Handu particularly inspiring is how it has turned traditional retail hierarchy on its head: instead of decisions flowing down from fashion directors and buyers, each micro-team acts as its own fashion brand, directly accountable to customer preferences. This structure has enabled them to launch new products in days rather than months, with each team learning and adapting through daily performance feedback.

Perhaps the most compelling example is Haier's RenDanHeYi ("people, needs, connections") model, developed under CEO Zhang Ruimin. This revolutionary approach has transformed the traditional corporate structure into over 4,000 micro-enterprises, each functioning as an independent profit and loss centre.

The Chinese saying, "It's better to be the head of a chicken than the tail of a phoenix," captures the cultural resonance of this approach, where employees prefer leading small, autonomous units to being minor players in larger hierarchies.

Alternative Models of Decentralisation

Beyond these corporate examples, we're seeing fascinating experiments in decentralised organisation across different sectors.

MakerDAO, the community-driven organisation behind the DAI stablecoin, operates without a central authority, with token holders around the world making key decisions on how the system evolves. Not without its challenges, it's a compelling example of how decentralisation can create transparency and resilience at scale.

The Mondragon Corporation in Spain demonstrates another powerful approach to decentralisation. As a federation of worker-owned cooperatives, Mondragon empowers teams to operate autonomously while collaborating within a shared purpose.

In the technology sector, companies like Gore-Tex and Valve Corporation have shown how decentralisation can drive creativity and engagement. Gore-Tex uses a "lattice" structure, where leadership emerges organically, while Valve's famously flat organisation gives employees the freedom to self-organise around meaningful projects.

These examples remind us that decentralisation isn't just about technology—it's about rethinking how we organise work, empower people, and make decisions in a way that's adaptive, innovative, and resilient.

What can we learn from these examples to prepare us for AI–enabled organisations?

What these pioneering organisations demonstrate is that decentralisation works best when combined with new mechanisms for coordination and support. Buurtzorg's digital platform, Haier's microenterprise system, and MakerDAO's governance protocols all show how technology can reimagine traditional management layers not by controlling employees, but by connecting them to the resources and information they need to make good decisions.

This is particularly relevant as we consider AI's role in organisations: perhaps instead of automating management decisions, AI could focus on enabling human decision-making at the edges of the organisation, much like these examples but at an even greater scale and speed – and even help employees collaborate more effectively.

Leadership in an AI-Enhanced Organisation

Looking at these pioneering organisations - from Haier's micro-enterprises to MakerDAO's community governance - we can see how both leadership and organisational structure need to evolve when power and capability become more distributed. The lessons point to three key priorities:

Decision Velocity: Organisations need frameworks that enable autonomous teams to act quickly while maintaining strategic alignment. Rather than navigating layers of bureaucracy, teams should be able to access resources and make decisions through AI-enabled platforms that understand both corporate constraints and market opportunities.

Customer Proximity: Following successful examples from Chinese organisations, teams should be structured around customer needs rather than internal hierarchies. Leaders need to focus on creating the conditions where entrepreneurial units can respond directly to market opportunities, supported by systems that help them access institutional knowledge and capabilities.

Resource Orchestration: Perhaps the most crucial leadership challenge is redesigning how teams access organisational resources. Instead of controlling and directing, leaders need to create environments where AI can serve as an intelligent mediator, helping front-line teams quickly identify and leverage relevant capabilities - from manufacturing capacity to intellectual property - without requiring constant permission-seeking.

The Future of Power Distribution

The challenge ahead lies not just in accessing expertise—AI is increasingly solving that problem—but in redesigning organisational structures to leverage both human and artificial intelligence effectively. The organisations that thrive will likely be those that can maintain the delicate balance between empowered front-line teams, AI-enabled coordination, and clear strategic direction – and to do that in a way that respects the culture and ways of working that are already established.

This evolution suggests a future where power flows not from positional authority or accumulated knowledge, but from the ability to orchestrate human insight, artificial intelligence, and organisational resources in service of customer needs. The middle layer of tomorrow's organisation might shift significantly, integrating intelligent systems to support faster, more informed, and more autonomous decision-making at every level, while enhancing human roles rather than replacing them.