Glossary
S
29 entries beginning with S, the same definitions the articles use.
Second-order effects
Indirect benefits that emerge later, such as faster processing raising customer satisfaction and reducing churn, or staff freed from repetitive tasks retraining into higher-value roles.
Related: Transforming the Board: Using Decision Analytics for Strategic Advantage
Selective Atrophy Concept
The intentional acceptance that some skills, like routine verification, may safely decline, while strategic reasoning, relationship building and complex judgement must be deliberately preserved through continued human engagement.
First introduced in: The Redeployment Dividend: Why AI Will Unleash Your People, Not Replace Them
Related: The Redeployment Dividend: Why AI Will Unleash Your People, Not Replace Them
See also:Cognitive atrophyIrony of automationRedeployment Dividend
self-supervised learning
Training on vast streams of unlabelled data such as video, sensor, and transaction feeds, so models learn underlying principles rather than memorising labelled examples, reducing dependence on expensive human labelling.
Related: World Models: The Next Horizon in AI for Predictive Enterprise Intelligence
See also:TrainingSupervised learning
Shadow AI
Employees’ use of unapproved AI tools outside enterprise controls, often successful informal pilots, revealing governance gaps rather than technology limitations.
Related: AI is transforming governance: Six key Boardroom prioritiesShadow AI amnesty: from discovery to governanceFrom Shadow AI to Strategic Asset: Building Your AI Centre of Excellence
See also:Shadow ITMinimum Lovable Governance
Shadow AI Discovery
An initiative that surfaces unofficial AI experiments through a no-penalty ‘register and receive support’ approach, anonymous surveys and procurement monitoring, turning hidden risk into a living, governed AI asset register.
Related: AI Centre of Excellence: Building Capabilities That Scale With AI Adoption
See also:Shadow AI
Shadow AI economy
The informal economy of personal AI tools employees use ahead of official programmes, productive but invisible to governance until it is deliberately surfaced.
Related: Crossing the GenAI Divide: Solving The 95% Problem With The Complete AI Framework
See also:Shadow AI
Shadow IT
Technology adopted outside the IT function’s sanction or oversight, historically unapproved software and cloud services, often ’not on the books’. The pattern that preceded shadow cloud and now Shadow AI: each wave shows demand outrunning governance, and each is answered with sanctioned alternatives rather than prohibition.
Related: The Board in the machine
See also:Shadow AI
Single point of failure
Single-model or single-jurisdiction dependence within a critical or regulated process, creating immediate, material continuity exposure if that capability is withdrawn, even by an order not aimed at the organisation.
Related: The AI Sovereignty Trilemma: When a Frontier Model Vanishes and Reality Bites
Six Board Concerns Asset
An interconnected lens of six concerns, Strategic Alignment, Ethical and Legal Responsibility, Financial and Operational Impact, Risk Management, Stakeholder Confidence and Safeguarding Innovation, that must be orchestrated together so AI discussion does not collapse into risk management alone.
Related: Rethinking Business Cases in the Age of AI: and Securing Buy-In from the BoardAI's Interconnected Challenge: Diagnosing the Six Concerns of the Board
See also:Minimum Lovable GovernanceAlignment
Small Language Model
A compact, often domain-specific AI model offering faster inference, lower cost, easier deployment, and reduced energy footprint, trading broad capability for higher accuracy within defined domains.
Related: Return-to-Work Briefing: Five Forces Reshaping the Board AI Agenda in 2026
See also:Inference
Small modular reactor
A smaller, factory-built nuclear reactor seen as a potential source of dedicated AI data-centre power, but with deployment timelines stretching into the 2030s, lagging behind AI’s accelerating energy demand.
Related: Beyond Regulatory Uncertainty: Thoughts on the UK's AI Sovereignty Challenge
Sovereign AI
AI capability a nation or organisation controls end to end, from infrastructure and data to models and their values, pursued to reduce dependence on foreign providers; the trade-offs are cost, capability lag, and talent.
See also:AI Sovereignty TrilemmaSovereign ControlOpen-weight
Sovereign Control Concept
The ability to keep an AI capability running on terms the organisation sets, rather than terms a distant provider, or a government, can revise without consultation; more than data residency.
First introduced in: The AI Sovereignty Trilemma: When a Frontier Model Vanishes and Reality Bites
Related: The Headroom Argument: Why AI Efficiency Means More Compute, Not Less
See also:AI Sovereignty Trilemma
Sovereign Specialisation Concept
A stance focusing entirely within one sovereignty domain, sacrificing global scale for deep alignment, clear governance, consistent stakeholder expectations and regional dominance.
First introduced in: AI Sovereignty: A Board's Guide to Navigating Conflicting National Agendas
See also:AlignmentAI Sovereignty TrilemmaAdaptive Localisation
Sovereignty Paradox Concept
The bind in which high domestic energy costs push UK businesses toward foreign-hosted AI services, creating strategic dependencies that undermine national AI-first ambitions even as they reduce immediate operational cost.
First introduced in: Beyond Regulatory Uncertainty: Thoughts on the UK's AI Sovereignty Challenge
See also:AI Sovereignty Trilemma
Sovereignty Sensing Concept
The organisational ability to detect early signals of regulatory shifts, infrastructure constraints or competitive positioning, giving decisive advantage over organisations that merely react to changes.
First introduced in: AI Sovereignty: A Board's Guide to Navigating Conflicting National Agendas
See also:AI Sovereignty Trilemma
Sparse attention
An architectural technique, exemplified by the SubQ launch, that reduces attention compute at full context by roughly 1,000 times, releasing previously uneconomic workloads rather than simply cutting cost.
Related: The Headroom Argument: Why AI Efficiency Means More Compute, Not Less
Stop Keep Change Model Asset
The verdict set of the Remake framework: for the work under examination the framework returns one of three verdicts, Stop, Keep, or Change, and all three are governed. A Stop verdict requires a decommission plan and verification that the stopping actually happened, a Keep verdict requires documentation sufficient to break key-person dependency, and a Change verdict proceeds through the full ADAPT motion. A remaking comprises all three, and an all-Keep outcome is a legitimate result.
See also:Process AuditADAPTRemake
Strategic augmentation
Using AI to support junior staff with proper frameworks so they learn faster, preserving expertise pipelines and gaining efficiency, contrasted with pure replacement that destroys future senior talent.
Related: The Accountability Gap: When AI Delegation Meets Human Responsibility
See also:expertise pipeline
Summary Substitution Failure Concept
When directors rely on an AI-generated summary of a board pack as their primary reading, applying statutory judgement to material whose framing and omissions the AI, not a human, chose.
First introduced in: AI and the Company Secretary: Operating the Boundary the Chair Polices
Superintelligence
A hypothetical AI substantially surpassing the best human minds across virtually all domains; a central concern of long-term AI safety debate.
See also:AI safety
Supervised learning
Training a model on labelled examples (inputs paired with correct answers) so it learns to predict the label for new, unseen inputs.
Related: World Models: The Next Horizon in AI for Predictive Enterprise Intelligence
See also:Training
Synthetic biology
Engineering biology by designing genetic circuits, organisms, and biological processes to specification; AI-driven design tools are collapsing its costs, moving programmable biology from research novelty toward industrial and security significance.
Synthetic data
Artificially generated data mirroring the statistical properties of real data, used to train AI models, test rare or risky scenarios, protect privacy and overcome data scarcity.
Related: Introducing the AI Stages of Adoption: A framework for understanding AI readiness in your business
See also:TrainingMachine learning
System prompt
A standing instruction that shapes a model’s tone, refusal posture and framing at the surface; it steers trained dispositions but does not rewrite them.
Related: Ethical AI: When the Model Imposes Values Your Organisation Did Not Choose
Systemic risk
Under the Code, a classification applied to frontier models exceeding 10^25 FLOPS, triggering risk assessments, adversarial testing, incident reporting to the EU AI Office, and cybersecurity measures.
Related: Navigating the AI Regulatory Maze: A Boardroom Survival Guide
See also:Frontier model
Systems thinking
The principle that optimising individual components of a system in isolation often produces worse outcomes than redesigning the whole; it is why AI workflow redesign compounds advantage over bolt-on augmentation.
Related: The Great Remaking: AI and the Race to Transform the Very Essence of Work
See also:Six Board Concerns
No S entries match. or try the site search.