AI Will Not Save a Weak Strategy - Rare Black Strategy and the Architecture of the Exponential Age
- Rare Writer
- 20 hours ago
- 7 min read
Artificial intelligence does not make an organisation strategic. It makes the organisation’s existing way of thinking faster, cheaper, more persistent and more scalable. That distinction is existential because many organisations have never clearly defined what strategy is.
A plan is not strategy nor is it a bunch of steps. A roadmap is not strategy. A framework, is not a strategy. Neither is a budget, technology programme, collection of priorities or set of targets. These are instruments operating inside an environment whose purpose, assumptions and boundaries have usually already been accepted.
From Rare’s triple-loop perspective -
Strategy is the disciplined capacity to perceive, question and reimagine the entire environment of concern—its purpose, actors, beliefs, assumptions, incentives, behaviours, systems, constraints and emerging possibilities—and then shape those conditions so better outcomes become likely before the existing game, and those within it, can dictate them or even know about them at all
Single-loop strategy asks whether the organisation is executing the existing game correctly. Double-loop strategy asks whether the plan, assumptions or even the game remain correct. Triple-loop strategy goes deeper: why does this game exist in its present form, which beliefs and behaviours reproduce it, who benefits from its architecture, and what entirely different environment could now be created?
At its highest level, strategy is not a response to the environment. It is the architecture, often invisible, of the environment itself.
Rare Black Strategy: the silent emergence
This is the territory of Rare Black Strategy—a low-signature, counter-anticipatory and emergent mode of strategy that quietly reconfigures the conditions of the game before its participants recognise that the game is changing.
It reshapes incentives, defaults, permissions, information, dependencies, decision pathways and strategic options. Harmful or obsolete behaviours become less effective. Constructive behaviours become easier, safer and more valuable. Participants need not experience a visible strategy programme; they experience altered conditions and a different distribution of opportunity, friction and consequence. By the time the strategic transition becomes visible, the former game may already be uneconomic, ineffective or impossible to play.
“Black” does not mean unlawful, deceptive or coercive. It means quiet, disciplined and low-signature. Its apparent invisibility comes from environmental design rather than covert manipulation; from preventing conflict rather than performing victory; and from changing the conditions around decisions rather than secretly controlling people.
AI becomes profound in this context. A model inherits the question it is asked. The question inherits the strategist’s assumptions. Those assumptions determine which reality the organisation can perceive. AI therefore amplifies both capability and limitation. It can industrialise insight—or industrialise blindness.
Strategic mode | AI typically explored | Strategic role of AI | Investment and environmental effect | Winner-Loser Position |
Transactional / incremental—reactive | Policy-first Generic copilots, chatbots, summarisation, content generation, RPA and basic workflow assistants. | Perform existing tasks faster, answer known questions and report what happened. | Tool licences, isolated pilots & basic productivity targets; workflows and decision rights remain largely unchanged. | Short-term efficiency with little differentiation. The loser automates inherited constraints and mistakes speed for transformation. |
Diagnostic | Process mining, document intelligence, retrieval systems, anomaly detection, causal analysis, knowledge graphs and evidence mapping. | Explain why outcomes occurred; connect records, behaviour and process reality; expose contradictions and causes. | Governed organisational data and workflow truth, tested through human judgement. | Better decisions and system repair. Advantage remains limited if diagnosis repeatedly restores a model that should be replaced. |
Predictive | Forecasting models, machine learning, propensity engines, streaming analytics, multimodal sensing, digital twins and scenario simulation. | Detect weak signals and model demand, risk, behaviour, capacity and environmental movement. | Context-rich data, leading indicators, continuous feedback, calibration and intervention thresholds. | Creates time advantage. Winners move before pressure becomes failure; losers forecast accurately but cannot convert foresight into action. |
Counter-anticipatory / Rare Black | Adversarial AI, red-team agents, multi-agent simulation, attack-path modelling, prescriptive optimisation and counterfactual engines. | Think from the future position of a competitor, threat or disruptive force; model moves and countermoves. | Reshapes incentives, controls, dependencies and strategic options before the anticipated move occurs. | Game-changing advantage. The anticipated move becomes ineffective, visible, uneconomic or impossible before participants recognise the terrain has changed. |
Emergent / silent emergence | Persistent contextual agents, multi-agent orchestration, adaptive digital twins, real-time environmental sensing, synthetic intelligence teams and governed adaptive learning | Discover what the situation is becoming; revise hypotheses, boundaries and objectives; generate and test new operating models. | A living human–AI environment connecting evidence, strategy, workflow, governance and execution. Investment centres on learning, option creation and reinvention | The winner does not merely play ahead—it helps create the next game. Laggards notice only after their former advantages become liabilities. |
This is not a simple technology ladder. An organisation can buy agentic AI and still use it transactionally. It can deploy advanced prediction to defend an obsolete annual plan. Conversely, relatively simple AI can support sophisticated strategy when embedded in strong human judgement and a wider intelligence environment.
The determining factor is not what the AI is. It is what strategic mode governs its purpose.
Activity is abundant. Transformation is scarce.
The market evidence already exposes the divide. McKinsey’s 2025 global survey found that b% of organisations used AI in at least one function, yet nearly two-thirds had not begun scaling it enterprise-wide and only 39% reported enterprise-level EBIT impact.
AI high performers—around 6% of respondents—look materially different. They pursue growth and innovation, redesign workflows, scale agents faster and demonstrate stronger leadership ownership. They are nearly three times more likely to redesign workflows, while approximately three-quarters are scaling or have scaled AI, compared with one-third of other organisations.
McKinsey’s 2026 readiness research found that leaders were 5.3 times more likely to report enterprise value where workflows had been redesigned: 32% versus 6% where they remained unchanged.
Indeed, BCG reports that leading companies allocate more than 80% of AI investment to reshaping key functions and inventing offerings rather than minor productivity initiatives. They focus on an average of 3.5 priority use cases, compared with 6.1 among others, and expect 2.1 times greater ROI. BCG’s wider research found only 5% of firms were “future-built”, while 60% received little material AI value.
The winners are not simply buying more intelligence. They are investing at a different strategic altitude.
The winners and losers
The losers will distribute horizontal AI, celebrate login rates, count generated documents and remove minutes from existing tasks. They may become impressively efficient while their value proposition, structure and decision model decay. Their AI will answer yesterday’s questions with extraordinary fluency.
The winners though, will connect AI vertically and environmentally to proprietary context, external signals, workflows, relationships, risks, behaviours and strategic objectives. They will ask not merely:
“How can AI perform this task?”
They will ask:
“Why does this task exist? What system reproduces it? What future is forming around it? Who could exploit it? What different environment would make it unnecessary—and should this organisation continue in its present form?”
Traditional planning treats the environment as something outside the organisation to which it must react. Triple-loop strategy recognises that organisations also produce environments themselves. They shape incentives, expectations, dependencies, information, categories and behaviour. They help determine which futures become easier or harder for every participant to pursue. Strategy at its highest level is therefore not prediction. It is the architecture of possibility.
Counter-anticipatory strategy changes the entire terrain before the threat, competitor or systemic pressure can exploit it. Emergent strategy recognises when the original problem, boundary or objective is dissolving and a different configuration has become possible or necessary or both. Rare Black-Strategy joins them to define and create this. It causes the new environment to emerge so coherently and quietly that participants adapt to a reality whose strategic architecture they may never have seen being constructed.
Shaking the biscuit tin
Yet many AI programmes resemble a board shaking an empty biscuit tin and demanding that a cake appear before the flour, eggs and butter have even left the supermarket shelf. The noise is reported as momentum. The shaking receives a maturity score. The dented tin enters the transformation register. Consultants debate whether it should be shaken in the cloud, with Cyber and AI mistakenly left to be governed in or from the boardroom. But there is still no cake as the board doesn't go to the store to get the ingredients let alone stir the mixture.
AI itself is the oven’s potential—not the ingredients, recipe, baker, kitchen or decision about what should be baked. Without coherent data, formed intelligence, environment-context mapped and understood in realtime, redesigned workflows, authority, governance, human capability and genuine strategy, more vigorous shaking merely produces louder evidence of absence.
This is why Rare Strategy brings strategy, digital discovery, systems quality, workflow truth, security, vendor control, organisational intelligence and AI into one executive environment. Rare i-Glue connects the evidence and operating layers. Triple-loop discipline tests not merely whether the plan is being executed, or whether the plan is wrong, but whether the organisation’s entire way of perceiving and shaping reality remains fit.
Rare Black-Strategy then, takes the decisive next step: shaping the environment before the environment dictates the organisation’s or situations fate. To know the environment as it needs to be is key to emergent Rare "Black" Strategy.
The winners of the exponential age will therefore not be those who merely purchase the strongest chess engine, deploy the most digital grandmasters or calculate the greatest number of moves. They will understand that strategy is not simply choosing the best move on the board placed before them.
While others use AI to optimise openings, protect pieces and calculate fifty moves ahead, the true strategist studies the game behind the game: who designed the board, why the pieces move as they do, which behaviours the rules produce, where every player is being drawn—and how those conditions can be re-architected before the clock even starts.
Within that deeper game, even the appearance of losing may be an intentional positional sacrifice: not a defeat within the existing contest, but part of reconfiguring the environment in which the eventual outcome will be decided. The highest strategic victory is achieved when the opponent’s strongest line becomes irrelevant and success emerges from the architecture of the board itself—not merely from the brilliance of the next move.
These strategists do not merely outplay the opponent. They quietly create a position in which the opponent’s strongest move loses its relevance, the old lines of attack disappear and the decisive outcome begins to emerge before anyone realises the original game has ended.
Everyone else may still possess extraordinarily powerful AI—a grandmaster whispering flawless moves into their ear. But they will be playing brilliant chess on an obsolete board, advancing with machine-perfect precision towards a checkmate architected by somebody else—before they have even understood which colour they are, whose game they entered, or that their future position was decided before their first piece moved.
Comments