The waggle dance is a story - a scout telling other bees what it’s found. Stories are the hive’s intelligence system. Here are stories everyone is now hearing. Names have been changed.
Operations director, Tjerk, sees rapid implementation of AI across his organization. He has an idea that the executive board might benefit. Could AI consolidate and report operational performance, capacity utilization, demand forecasts, inventory levels, contingent staffing needs, supplier performance and risk indices?
He told Thomas, Head of IT, to look into it. How hard could it be? All around the organization, there was daily new evidence of AI adoption at whiplash rates.
Suzanne, Global Supply Chain Performance Manager in London, is approaching retirement. After paying EUR 4000 for professional tax guidance, she got identical advice from an AI.
An excited Marcus in HR brought Tjerk the results of a market-research assignment given to a job applicant. The candidate had used AI to forecast the strategies of the organization’s major competitors. Tjerk struggled to hide his shock at the insights.
Michaela, executive assistant to the Head of Finance in Paris, is relieved to be no longer responsible for knowledge management. Notes and reports now go through unfiltered to everyone in the room – and to some who were not.
Maria is an ambitious junior lawyer in Milan. She had spent most of her time on document review. With AI-automation she now deals only with those flagged as uncertain; and is building a different, and sharper, judgment on risk than any predecessor.
Gasim in the Omani Logistics Centre monitors global freight. Previously he would escalate half of all non-conforming movements. With AI support, he now deals with 90% of exceptions himself.
The CEO was quietly proud of his son, Paul. Formerly a code-writer, Paul is now a solutions architect on five times the salary. His colleagues were laid off when code-writing went to AI. He is hoping to get into Tech Leadership before AI reduces the need for solutions architects.
Tjerk tries to make sense of it. People adapt – they always do – but he knows this is different. It reminds him of the film, “Everything, Everywhere, All at Once”. Where is leadership in all of this?
Emergent Complexity
Tjerk had resolved it was time for the board to exploit AI and asked Thomas to launch a control-tower project.
Thomas was in Tjerk’s office with three external consultants. Tjerk wondered why Thomas needed support.
After hearing the pitch, Tjerk stared at Thomas for a moment. Then...
“Let me check”, Tjerk said. “Are we, or are we not, in the second quarter of this century? Because your excuse is straight out of the last one. How is it possible we are still in a fog of bad data?”
Tjerk was surprised at his own waspishness.
The frustration is raw and has three parts: complexity, information, and leadership.
Complexity
Complexity is emerging at operational level. Staff are finding new opportunities for AI in work and at home. It catalyzes spontaneous, bottom-up change and gives staff agency. This drives heterogeneity. Intelligence is moving from the centre to the edge. Line-managers from a culture built on standardization and control are challenged.
Information
Information leakage has a new form, not only via transcription and chat-log apps. Companies radiate data by existing, ‘data-exhaust’. With AI help, third parties can infer a company’s operational and strategic intentions from it. Where is it coming from? Can it be controlled? What are competitors, analysts, investors, suppliers, customers and activists able to learn from data-exhaust?
Leadership
For decades, productivity has been leadership’s prime goal. But is it enough just to do the same things more efficiently? New competition, alternative solutions and changing requirements are not new. What has changed? AI has shrunk the time from first awareness to critical action needed. The question that stands above others is:
How does the board learn to look outward — before it is too late to matter?
Externalities
Today’s business leaders have seen relative stability in business, geopolitics and society. Success was productivity-driven gains – faster, cheaper, better. Three decades have passed since publication of “The Innovator’s Dilemma: When New Technologies Cause Great Firms to Fail” by Clayton Christensen. The management trend since then has been ‘digitization of the status quo’. The conclusion in a Sandler report by the Ruby Group “What CEOs actually want” (December 2025) reads:
B2B CEOs entering 2026 are focused on:
Predictable revenue generation
Scalable systems and repeatable processes
Sales performance and pipeline execution
Leadership development and talent readiness
Measurable business outcomes and ROI
These are internal productivity metrics. They live in scientific management and balanced score cards. The domain of supervisors and managers. Not leaders. CEOs who prioritize them are indistinguishable from managers.
Executives with this mindset see the emergent complexity from AI as a threat or ‘governance stress test’ to their administrative order. They turn to internal guardrails, ‘plumbing’ and workflow-policing; not scanning the horizon for threats and opportunities.
A productivity mindset was right for some and for its time.
In a stable environment, it is expensive to commit resources to horizon-scanning; even more so to maintaining a capability to respond to events that are infrequent and slow-moving. The organization handles them with existing resources - after they have optimized present operations. It does not matter that external awareness gets a lower priority. Horizon-scanning is genuinely difficult. What does it even look like? How do you measure it? How do you know when it has worked? Its outputs are qualitative, probabilistic and often unwelcome. It is easier to launch a control-tower than an activity whose product is “we think something might happen“. It seems ethereal and distracting. External scanning and evaluation was firstly de-prioritized, then removed. The reasons are easily understood and most organizations do not have it today. That can be the right thing to do – maintain an inward, productivity focus – in a stable environment.
The environment is no longer stable. This is not a temporary condition. Complexity and intelligence is emerging at the edges. Experience in optimizing complicated systems is inadequate for dealing with complexity. The formerly predictable ‘five forces’1 (supplier power, competitive rivalry, substitution threat, buyer power, new entrants) are AI-turbocharged. The acceleration of AI-driven change means that the gap between a threat first being visible and becoming critical is shrinking. Reactive response, which worked when threats moved slowly, is no longer adequate.
The external overview of your own organization is now available to everyone. A potential acquirer, an activist investor, an aggressive competitor can now generate a sophisticated, strategic picture of your organization in hours. Without external vision, your organization is not just impaired — it is so in an environment where others have night-vision goggles and first-person-view drones.
Described formally in ‘The Brain of the Firm’ by Stafford Beer2 in 1972, external overview capability (System 4 in Beer’s Viable Systems Model) was never easily acquired. Continuous environmental scanning needs analytical capacity, pattern recognition across disparate domains, and synthesis of weak signals into coherent models. That was a long way off in 1972 but, despite the difficulties, many organizations before 2000 had corporate planning departments responsible for long-range vision and scenario planning. That changes sometime around 2000. McKinsey’s ‘Corporate Horizon Index’ reports a compression of time horizons as companies replaced human insight with risk and stress-testing metrics on current operations.
AI makes the external overview easier to create. The reason it was hard — continuous environmental scanning requires enormous analytical, inference and predictive capability— is why AI is good at it. The capability that Beer described as necessary, but left organizations to struggle with, is technically accessible for the first time.
The capability cannot be purchased as a product It cannot be implemented as a project. I t is not a reporting tool. It is an organizational capability — a way of thinking, asking questions, and acting on uncertain and incomplete information about the future. AI makes it feasible, but only if the organization understands what it is and resists the temptation to treat it as a technology project. It requires people who are oriented outward, who are protected from the distraction of current operations, and who have the authority to bring unwelcome findings to the board.
AI supports that capability. It does not replace it. Organizations that commission an AI to do their environmental scanning – and believe the output – will get it wrong.
When a beehive is well supplied with nectar, scouting activity declines. Interest in the scouts’ stories also declines and fewer bees set off to explore the new nectar sources. Stories in corridors, cafeterias and conferences are the waggle-dance. Who is listening? How are they understood? And who has the standing to tell the board what is out there?
1Porter’s Five Forces - https://en.wikipedia.org/wiki/Porter%27s_five_forces_analysis
2Anthony Stafford Beer - https://en.wikipedia.org/wiki/Stafford_Beer
Sharp piece. Following. From what I've seen.. the firms that get blindsided are the ones where scanning got folded into strategy offsites twice a year.
You know... episodic, not continuous. The real bottleneck was never the AI. Thanks for pulling this together!
Thank you for sharing this perspective. It is sharp and clearly reflects the situation I am increasingly hearing as well.
Your point on data exhaust is particularly urgent. Competitive intelligence practices now connect signals from vast amounts of uncontrolled information, making it possible to infer a company’s strategic direction, often without realizing how much it is revealing.