Always Catching Up on AI? It's Not About the Tools | Connexxo

Requisite Variety and the Adaptive Organisation in the AI Era

This article is based on a talk given at the AI Transformation Camp Meetup. It summarises the central argument: the organisations that will thrive with AI are not those that adopt the most tools fastest, but those that build the capacity to keep adapting.

Many leaders feel permanently behind on AI. A new model, a new agent, a new tool appears every week, and the instinct is to chase it. That instinct is understandable, and it is also misdirected. The enduring advantage in the AI era does not come from the tools. It comes from the kind of organisation you are when the tools arrive. This article explains why, using a law from cybernetics that is more than half a century old and more relevant now than ever.

AI as an Amplifier

The most important thing to understand about AI is that it does not transform organisations uniformly. It amplifies them. Introduce powerful AI tooling into an organisation with strong foundations, clear feedback loops, disciplined quality practices, and psychological safety, and you get amplified output and compounding gains. Introduce the same tooling into an organisation with weak foundations, and you amplify the dysfunction and accelerate the failure.

Figure 1. AI amplifies the direction the organisation is already travelling in. Good organisations become better; struggling organisations become worse.

The point lands more sharply with a concrete example. Ask yourself what happens to your code reviews if your organisation suddenly produces ten times as many lines of code as it does today. The bottleneck does not disappear; it moves, and it moves to whichever part of your system was already the weakest. If review capacity, quality discipline, or shared understanding were thin before, AI does not fix them. It floods them.

There is a second amplification at work, and it operates on the outside of the organisation rather than the inside. AI increases the dynamics of markets. Competitors move faster, customer expectations shift faster, and the ground underneath any given strategy moves faster. You will have to navigate in even more disturbed waters than before. This is where a deeper question arises: what determines whether an organisation can cope with a more turbulent environment at all?

Requisite Variety: The Core Idea

The answer was formulated in 1956 by the cybernetician W. Ross Ashby, in his Introduction to Cybernetics. It is known as the Law of Requisite Variety, and Ashby expressed it like this:

The variety of the regulator must be at least equal to the variety of the disturbances to be attenuated.

Variety here is a precise term. It is the number of distinct states a system can be in. In formal terms, if the environment can generate n distinct states, the regulator needs at least n distinct responses, or some of those disturbances will inevitably pass through uncontrolled. Only variety can absorb variety. A system can only regulate what it has enough internal range to respond to.

Requisite Variety Applied to Organisations

The law transfers directly to organisations, and the translation is what makes it powerful.

The environment is the source of variety. Markets shift, competitors act, regulations change, and customers behave unpredictably. Each distinct state the environment can be in represents a unit of variety the organisation must absorb.

The organisation is the regulator. Its structure, its decision-making capacity, its skills, and its repertoire of responses constitute its own variety.

The law then makes a stark claim. If your organisation has less internal variety than the environment it operates in, it will lose control. It cannot respond to disturbances it has no response for. This is not a matter of effort or intention. It is a structural fact.

Figure 2. Requisite Variety for organisations: only variety can absorb variety. The organisation stays in control when its variety matches or exceeds its environment's.

Requisite Variety in Practice

The law shows up in recognisable, everyday organisational patterns.

Too little internal variety. A rigid hierarchy that can only respond to problems in pre-approved ways. Novel threats are met with the same old playbook, and the organisation is surprised again and again by situations it has no response for.

Variety mismatch in structure. A centralised function trying to manage a globally diverse operation. The centre lacks the local knowledge and the speed to match what the periphery actually faces, so its responses arrive late, generic, and wrong for the situation.

Amplifying variety. Decentralisation, cross-functional teams, and psychological safety all increase the organisation’s response repertoire, raising its variety to match the environment. And AI, used well, belongs on this list too: it can extend the range of responses an organisation can generate.

Reducing environmental variety. Standardisation, contracts, and regulations reduce the number of states the environment can present. This is a legitimate strategy that lowers the bar the organisation has to reach, rather than raising the organisation to meet it.

The Two Adaptive Strategic Responses

Because the law requires variety to match, there are only two levers for adaptivity.

Amplify internal variety. Hire diverse talent, decentralise authority, build adaptive capacity, and tolerate experimentation. This raises the organisation up to the level of its environment’s variety.
Attenuate external variety. Simplify the product range, lock in long-term contracts, lobby for regulation, and focus the market segment. This lowers the environment down to a level the variety of the organisation can manage.

Most organisations need both at the same time. The two levers are not in opposition; they are complementary moves in the same balancing act, both working towards the same goal: matching the organisation’s variety to its environment’s.

Figure 3. The two strategic responses. Most organisations need both simultaneously.

Why Most Hierarchies Fail This Test

Traditional command-and-control structures deliberately destroy variety. Standardised processes, narrow job roles, and approvals flowing upward are all mechanisms for reducing the number of distinct states the organisation can be in. This is efficient when the environment is stable and predictable, and for a long time it was the right design.

When the environment becomes turbulent, through technological disruption, geopolitical shocks, or rapid market change, the organisation’s variety drops far below what is required, and control breaks down. The law is unforgiving: you cannot wish, mandate, or reorganise your way around it. If the variety is not there, the control is not there. AI, by increasing the dynamics of the environment, pushes more organisations across this threshold faster.

Adaptive Capacity: The Real Differentiator

This brings us to the central claim. The organisations that will compound their advantage over time are not those that acquire the best AI tools fastest. They are those that build the adaptive capacity to keep learning and adjusting as AI continues to evolve.

AI capabilities are becoming table stakes. Every serious organisation will reach a functional level of AI-augmented work within a few years. The competitive question is not who gets there first. It is who builds the organisational capacity to keep adapting after they get there, and after the tools change again. That capacity has five dimensions that consistently distinguish organisations that sustain advantage from those that achieve a moment of advantage and then plateau.

Solution orientation. Genuine, ongoing focus on customer and user outcomes, not on tool capabilities as goals in themselves. Organisations oriented to solutions keep asking whether AI is producing better outcomes for real people. Organisations oriented to capabilities keep acquiring new tools and measuring their own sophistication.

Fast improvement cycles. The organisational discipline of learning from experience quickly and adjusting deliberately. In the AI context, this means running experiments with honest measurement, capturing what works, intentionally spreading it, and not waiting for perfect evidence before updating practices.

Organisational multidexterity. The structural ability to operate at three tempos simultaneously: stable operations that keep current systems reliable, capability development that builds the competence to operate at higher levels of AI integration, and horizon scanning that watches where the market and AI are heading and makes early bets on what matters. All three have to happen in parallel, and in synergy. Organisations that collapse these tempos, letting the urgent crowd out the important or letting the experimental destabilise the operational, lose the capacity to adapt before they lose their competitive position.

Figure 4. Organisational multidexterity: three tempos sustained at once, rather than one at a time.

Adaptive leadership. Leaders who treat uncertainty as the permanent operating condition rather than a temporary disruption, who invest in learning infrastructure rather than directing from certainty, and who protect the human foundations of psychological safety, cognitive sustainability, and the talent pipeline on which AI quality ultimately depends.

Healthy relationships. The trust, psychological safety, and open communication that make honest learning possible. An organisation in which a majority of employees conceal their use of AI cannot learn from its own experience. The relational substrate is not a soft concern in the AI transformation; it is the medium through which organisational learning travels.

Figure 5. The five dimensions of adaptive capacity, seen as a radar profile. The shaded area is the organisation's adaptive capacity: a weak dimension shrinks the area, so the system is only as strong as its smallest dimension.

These five dimensions are not independent. They are a system. When any one is absent, the others are weakened. When all five are present and sustained, the organisation compounds its advantage: each new wave of AI capability finds a team that is better able to use it well, learns from it faster, and adjusts more deliberately than the time before.

Figure 6. Adaptive capacity compounds: each wave of AI finds a more capable organisation than the one before.

Conclusion

If you feel you are always catching up on AI, the remedy is not to run faster after the tools. The tools will keep changing, and chasing them is a race with no finish line. Requisite Variety tells us what actually determines control in a turbulent environment: an organisation needs as much internal variety as the world throws at it. AI raises the variety of the environment, so it raises the bar. The organisations that clear it are the ones that have built adaptive capacity, the standing ability to keep learning and adjusting. That, and not the latest tool, is the defining capability of the AI era.

Dr. Jürgen Hoffmann is an Enterprise Coach and Trainer who holds a PhD in semiconductor physics and pioneered Scrum adoption in 2004, guiding agile transformations across the automotive, energy, finance, and technology sectors. He is the author of two books on agile methods and a Qualified Educator of the Society for Adaptive Organisations.

Pierluigi Pugliese is an experienced agile coach, systemic consultant, and trainer with over thirty years of practical experience in product development. Based in Munich and Italy and working through Connexxo GmbH, he supports large-scale agility, advises senior leadership, and is a Qualified Educator of the Society for Adaptive Organisations.

References

Ashby, W. R. (1956). An introduction to cybernetics. Chapman & Hall.
Beck, P., Demirel, I., Gadet, M., Hoffmann, J., Mueller, R., Pugliese, P., Roock, S., Schliep, A., Silva, R., & Zanotto, D. (forthcoming 2026). Adaptive organisations: Thriving in complexity. Society for Adaptive Organisations.