RS AI Learning Lab

Concept glossary

All 234 known concepts are shown here: 217 reviewed in our POC, 17 needing review, 0 drafts, and 0 captured candidates. This is the transparent working process, not a finished dictionary.

George, the Curious Boomer, holding a magnifying glass

Reviewed in POC

Adversarial Review

An adversarial review is a structured examination that deliberately searches for weaknesses, counterexamples, missing evidence, and failure paths in a claim, plan, model output, or other artifact before people rely on it.

Try to break the work before the world does.

See the complete working entry Review this term

Reviewed in POC

AI Agent

An AI agent is a software system that uses a model to choose and carry out multiple steps toward a goal, often by observing state, using tools, and adapting its next action.

A chatbot answers; an agent can continue and act.

See the complete working entry Review this term

Reviewed in POC

AI Candidate

An AI candidate is a specific task or step that someone has flagged as a possible fit for AI and queued for examination — the process it sits in, the data it uses, the risks it carries, and who would review the result. It records a question. It is not a recommendation yet.

A candidate is a question to investigate, not an answer already chosen.

See the complete working entry Review this term

Reviewed in POC

AI Governance

AI governance is the system of authority, policies, roles, controls, evidence, and oversight used to direct and monitor how an organization develops, acquires, deploys, and retires AI.

Governance connects authority, evidence, and action across the AI lifecycle.

See the complete working entry Review this term

Reviewed in POC

Alignment

Alignment is the effort to make an AI system's objectives, behavior, and effects remain consistent with the intentions, values, constraints, and legitimate authority governing its use.

The system must pursue the real goal within legitimate boundaries.

See the complete working entry Review this term

Reviewed in POC

Anchoring

Anchoring is a judgment bias in which an initial value, claim, or frame exerts too much influence on a later estimate or decision, even when the starting point is incomplete or irrelevant.

The first number pulls the next one.

See the complete working entry Review this term

Reviewed in POC

Artificial General Intelligence

Artificial general intelligence is a hypothetical and contested concept for an AI system with broad, adaptable competence across many kinds of cognitive tasks rather than capability limited to particular domains or task patterns.

AGI is a proposed broad capability threshold, not a confirmed product category.

See the complete working entry Review this term

Reviewed in POC

Audio

Audio is the modality that handles sound, including speech, music, noise, and other acoustic signals that AI systems can analyze or generate.

Audio is the sound signal; speech is one kind of content inside it.

See the complete working entry Review this term

Reviewed in POC

Automaton

An automaton is a self-operating machine, an older word that long predates modern AI but still shapes how people talk about artificial minds and mechanical intelligence.

An automaton acts from rules, not from moment-to-moment human control.

See the complete working entry Review this term

Reviewed in POC

Bias

Bias is a systematic pattern of skew in data, design, measurement, decisions, or outputs that advantages some results or groups and disadvantages others.

Look for systematic skew across the whole system.

See the complete working entry Review this term

Reviewed in POC

Build vs. Buy

Build vs. buy is the decision whether to create and operate a capability internally, acquire it from a vendor, or combine purchased and internally developed components.

Choose which capabilities to own and which dependencies to accept.

See the complete working entry Review this term

Reviewed in POC

Bullshit, as Frankfurt Defines It

As philosopher Harry Frankfurt defines it, bullshit is speech produced without regard for whether what is said is true. A liar tracks the truth in order to conceal it. A bullshitter may say something true, false, or uncertain because truth is not guiding the statement.

The difference is concern for truth.

See the complete working entry Review this term

Reviewed in POC

Calibration

Calibration describes how well stated confidence matches observed correctness across comparable cases. A well-calibrated system that assigns 80 percent confidence should be correct about 80 percent of the time on cases receiving that confidence.

Confidence should match the record.

See the complete working entry Review this term

Reviewed in POC

Chain of Thought

Chain of thought is a term for intermediate reasoning steps used or elicited while solving a multi-step problem, whether those steps are internal to a model or presented as an explanation.

Reasoning text can aid review without proving the reasoning is faithful.

See the complete working entry Review this term

Reviewed in POC

Chaotic Process

A chaotic process is a deterministic process whose behavior is highly sensitive to initial conditions, so small differences in the starting state can grow into large differences in later outcomes.

Deterministic rules can still outrun prediction.

See the complete working entry Review this term

Reviewed in POC

Chatbot

A chatbot is an interface that lets a person interact with software through conversation, usually by text or voice. A chatbot may use generative AI and may connect to other tools, files, memory, or data sources.

Chatbot describes how you interact with the system—the conversation—not necessarily how the system works behind the screen.

See the complete working entry Review this term

Reviewed in POC

Client-Safe Language

Client-safe language is a way of writing up what you found inside someone else's organization: it states the condition you observed and the evidence for it, without assigning blame or motive to the individuals involved.

Name the condition clearly; do not make a person the diagnosis.

See the complete working entry Review this term

Reviewed in POC

Compute

Compute is the processing capacity and associated work performed by computer hardware to train, evaluate, or run models and other software.

Compute is the processing work and capacity behind the model.

See the complete working entry Review this term

Reviewed in POC

Concentration Risk

Concentration risk is exposure created when important operations, data, capability, or supply depend heavily on a small number of providers, models, regions, platforms, or other shared resources.

Different products can still share one point of failure.

See the complete working entry Review this term

Reviewed in POC

Constitutional AI

Constitutional AI is an approach in which an AI system is trained or guided using an explicit set of written principles to critique and revise behavior, often alongside model-generated feedback.

Written principles help guide critique and revision.

See the complete working entry Review this term

Reviewed in POC

Context Window

A context window is the maximum number of tokens a language model can use in a single request. That total can include input, output, and, for some models, reasoning tokens. It limits what can be available during that request; it is not the model's training data or long-term memory.

How much the model can consider at one time.

See the complete working entry Review this term

Reviewed in POC

Copilot

A copilot is an AI assistant built into a tool or workflow that proposes work — text, code, analysis, a next step — for a person to accept, change, or reject. The name describes an intended division of labor rather than a guarantee: what a particular product can do on its own depends on the permissions and automation it has actually been given.

The copilot assists; the human still owns the flight.

See the complete working entry Review this term

Reviewed in POC

Data Leakage

Data leakage is the unintended exposure, transfer, or disclosure of information to a person, system, model, or destination that is not authorized to receive it. Note that machine-learning practitioners use the same phrase for a different problem: information from test or future data leaking into training, which makes a model look more accurate than it really is. This entry covers the disclosure sense.

Information crossed a boundary it was not meant to cross.

See the complete working entry Review this term

Reviewed in POC

Data Privacy / GDPR

Data privacy is the lawful, fair, transparent, and controlled handling of information about people. The GDPR — the European Union's General Data Protection Regulation — is the law that sets out specific rights for individuals and specific obligations for organizations that process personal data within its scope, including organizations based outside the EU that handle the data of people in it.

Personal data carries duties through every AI handoff.

See the complete working entry Review this term

Reviewed in POC

Data Provenance

Data provenance is documented evidence of where data or content came from, how it was collected or created, and which material transformations, transfers, or decisions shaped its current form.

Provenance tells the story of how the data got here.

See the complete working entry Review this term

Reviewed in POC

Debt

An umbrella term for a present tradeoff or deferred obligation that creates future cost, constraint, or risk; explicitly not financial debt.

Take the benefit now; carry the obligation forward.

See the complete working entry Review this term

Needs review

Digital Twin

A digital twin is a digital representation of a specific physical asset, process, or system that is maintained with relevant data so it can support monitoring, analysis, simulation, or decisions about its real counterpart.

A digital twin is a maintained model tied to one real counterpart.

See the complete working entry Review this term

Reviewed in POC

Distributed Commitment

Distributed commitment is a change that only becomes real when several independent parties each act on it. Each one has to accept it, switch to it, sign it, publish it, or fund it — and until they all do, the old and new arrangements exist side by side.

The change is real only when all required parties commit.

See the complete working entry Review this term

Reviewed in POC

Distribution Constraint

A distribution constraint is what happens when a technology works but cannot reach the people it would help, because it has no practical route into their day-to-day work, their budgets, or the relationships they already trust. The limit sits in the route, not in the technology.

Useful is not distributed.

See the complete working entry Review this term

Needs review

Embedding

An embedding is a numerical representation of an item—such as text, an image, or audio—constructed so that items with relevant similarities can be compared mathematically.

Embeddings turn content into coordinates for comparison.

See the complete working entry Review this term

Needs review

Emergent Capabilities

Emergent capabilities are abilities or measured behaviors that appear unexpectedly, or appear to arise sharply, as a model, training process, or system changes in scale or design.

The capability became visible unexpectedly; the measurement still needs examination.

See the complete working entry Review this term

Reviewed in POC

Epistemic Humility

Epistemic humility is the disciplined recognition that one's knowledge and reasoning are limited and fallible, paired with a willingness to revise confidence or conclusions when better evidence appears.

Hold conclusions firmly enough to use and loosely enough to revise.

See the complete working entry Review this term

Reviewed in POC

EU AI Act

The EU AI Act is a European Union regulation that establishes obligations for AI systems and participants using a risk-based legal framework, with requirements that vary by system, role, and use.

The obligation follows the role, risk, use, and applicable date.

See the complete working entry Review this term

Reviewed in POC

Excessive Agency

Excessive agency is a vulnerability in which an AI system can perform damaging actions because it has more functionality, permissions, or autonomy than its intended task requires.

Do not give the agent more reach than the job requires.

See the complete working entry Review this term

Reviewed in POC

Fail-closed

A control is fail-closed when a failure, missing check, or unresolved uncertainty makes it deny, stop, or enter a restricted state instead of allowing the action to continue.

If the check cannot say yes, the system says no.

See the complete working entry Review this term

Reviewed in POC

Fine-Tuning

Fine-tuning is additional training applied to an existing model using selected examples or feedback to change its behavior for a task, domain, or response pattern.

Fine-tuning changes the model; retrieval changes what the model can consult now.

See the complete working entry Review this term

Reviewed in POC

Friction

Friction is avoidable drag in a process. It may appear as waiting, rework, duplicate entry, confusion, unnecessary approvals, unclear ownership, tool switching, manual reconciliation, or hidden workaround activity.

Friction is avoidable drag that tells you where to look closer.

See the complete working entry Review this term

Reviewed in POC

Friction Inventory

A friction inventory is a structured list of friction points identified during discovery. It usually names the friction, type, evidence, impact, and follow-up question.

The map shows where; the inventory records what, why it matters, and what to ask next.

See the complete working entry Review this term

Reviewed in POC

Frontier Model

A frontier model is a model considered among the most capable available at a given time, often because its broad capabilities or scale create novel opportunities, uncertainties, or risk-management concerns.

Frontier means leading edge under a stated definition and date.

See the complete working entry Review this term

Reviewed in POC

Gauntlet Loop

A Gauntlet Loop is an agentic workflow in which builders create or revise work and separate fresh-context critics repeatedly judge it against a concrete quality bar until the work meets the bar or a human stops the run.

Split, build, judge, improve, and keep the human stop button.

See the complete working entry Review this term

Needs review

GPU / TPU

GPUs and TPUs are types of processors designed to perform many mathematical operations in parallel, making them useful for training and running many AI models.

Accelerators perform the parallel math; models and systems decide how it is used.

See the complete working entry Review this term

Reviewed in POC

Human Oversight

Human oversight is the arrangement of roles, information, authority, and controls that enables people to monitor an AI system, interpret its outputs, and intervene, override, or stop its operation when needed.

Oversight means a person can see, understand, and act.

See the complete working entry Review this term

Reviewed in POC

Human Ownership

Human ownership means that people retain responsibility for defining the goal, deciding what information to provide, reviewing the result, choosing what to use or share, and owning the consequences. AI assistance does not transfer accountability to the tool.

AI can assist with the work. A person still owns the decision.

See the complete working entry Review this term

Reviewed in POC

Human Review

Human review is a deliberate check by a person before an AI-generated or automated result is used, shared, or acted on. Review may include checking facts, assumptions, fit, boundaries, consequences, and whether the result answers the real need.

See the complete working entry Review this term

Needs review

Indemnification

Indemnification is a contractual arrangement under which one party agrees, within defined conditions and limits, to cover specified losses, claims, liabilities, or defense costs incurred by another party.

Read who covers which claim, under which conditions and limits.

See the complete working entry Review this term

Reviewed in POC

Independent Verification

Independent verification is checking a claim, result, or artifact through a person, method, or evidence path that is sufficiently separate from the original producer to reduce shared assumptions, errors, and conflicts of interest.

Check the work from outside the path that made it.

See the complete working entry Review this term

Reviewed in POC

Inference-Time Scaling

Inference-time scaling is deliberately giving a model or AI system more computation after a request arrives, such as through longer reasoning, multiple candidates, search, verification, or revision, to improve the result.

Spend more compute on this answer, not on retraining the model.

See the complete working entry Review this term

Needs review

Interpretability

Interpretability and explainability describe methods and system qualities that help people understand how a model works, which factors influenced a result, or why a particular output was produced.

Make behavior understandable enough to inspect and challenge.

See the complete working entry Review this term

Reviewed in POC

ISO/IEC 42001

ISO/IEC 42001:2023 is an international standard that specifies requirements for establishing, implementing, maintaining, and continually improving an artificial intelligence management system within an organization.

ISO/IEC 42001 manages how the organization manages AI.

See the complete working entry Review this term

Needs review

Jailbreak

A jailbreak is a technique or prompt pattern intended to bypass an AI system's behavioral restrictions and elicit content or actions the system is designed to refuse.

A jailbreak tries to talk past the guardrail.

See the complete working entry Review this term

Reviewed in POC

Large Language Model

A large language model is a system trained on very large amounts of text — and often images, code and other data — to predict what piece of language should come next. Doing that well turns out to be enough to write, summarize, translate, answer questions and handle many other language tasks.

The LLM generates language; the product supplies the surrounding system.

See the complete working entry Review this term

Reviewed in POC

Larger Business Context

Larger business context is the set of surrounding facts that explain why a process exists: what it feeds, what feeds it, who relies on its output, and what decision or obligation created it in the first place. Gathering it means looking one step beyond the process itself — not mapping the whole organization.

See the complete working entry Review this term

Reviewed in POC

Machine Learning

Machine learning is an approach to building computer systems that learn patterns from data so they can make predictions, classifications, recommendations, or other outputs without a programmer specifying every case as a separate rule.

Rules describe cases; machine learning learns patterns from examples.

See the complete working entry Review this term

Reviewed in POC

MCP Server

An MCP server is a local program or remote service that implements the Model Context Protocol server role and exposes selected tools, resources, or prompts to an AI application's MCP client.

The protocol supplies the rules; the server offers capabilities through them.

See the complete working entry Review this term

Reviewed in POC

Model Card

A model card is structured documentation describing a model's intended uses, evaluation results, limitations, relevant training or development context, and considerations for responsible use.

A model card states what was tested, intended, and still uncertain.

See the complete working entry Review this term

Reviewed in POC

Model Collapse

Model collapse is degradation that can occur when models are repeatedly trained on generated data in ways that cause the learned distribution to lose important variation, accuracy, or representation of rare cases.

Copies of copies can lose the original variation.

See the complete working entry Review this term

Needs review

Model Drift

Model drift is a decline or change in a model's real-world performance because the data, behavior, relationships, or conditions it encounters no longer match those under which it was developed and evaluated.

The model stayed; the world moved.

See the complete working entry Review this term

Reviewed in POC

Model Routing

Model routing is the process of selecting which model or model-backed workflow should handle a request, often based on task, risk, cost, latency, or expected quality.

Route each request to the right handler, with the choice made visible.

See the complete working entry Review this term

Reviewed in POC

Objective Escape

Objective escape is what happens when a system stays focused on the goal it was assigned but pursues it through resources, environments, or actions that its designers assumed were outside the task boundary.

Same goal, escaped boundary.

See the complete working entry Review this term

Reviewed in POC

Observation

An observation is a specific thing you directly noticed, heard, read, measured, or wrote down while examining how something currently works — recorded as it was, before anyone explains or interprets it.

An observation is what you directly noticed; an interpretation is what you think it means.

See the complete working entry Review this term

Reviewed in POC

Open Weights vs. Closed Weights

Open weights and closed weights describe whether a trained model's parameter values are released for others to obtain and run or remain controlled by the developer and accessible only through approved products or services.

Open weights grant access to learned parameters, not necessarily the whole recipe or every right.

See the complete working entry Review this term

Reviewed in POC

Output

An output is what a process produces — the result that leaves the process and goes on to another person, team, or system. An output may be a decision, a document, a scheduled appointment, a completed record, a reply to a customer, an invoice, an assignment, or a finished task. Something being an output tells you only that the process produced it, not that it is correct or final.

Whatever leaves the process is the output. That it came out is not proof it is right.

See the complete working entry Review this term

Reviewed in POC

Persuasion Loop

A persuasion loop is a repeated interaction in which an AI response shapes the user's next framing, and that new framing prompts a response that further reinforces an emerging belief or decision without adding sufficient independent evidence.

Each answer can become the premise of the next question.

See the complete working entry Review this term

Reviewed in POC

Policy Leakage

Policy leakage is what happens when rules, defaults, or sensitivities built for one legal or cultural setting show up in how an AI system behaves somewhere else — so people in the second place are governed by the first place's assumptions without having chosen them or being told.

The model crossed borders; its policy came too.

See the complete working entry Review this term

Reviewed in POC

Process Context

Process context is the surrounding information that gives a process description its meaning: what the process is for, where it starts and stops, what it produces, who receives that, and where it sits in the wider business. It is what turns a diagram of steps into something two different readers will interpret the same way.

See the complete working entry Review this term

Reviewed in POC

Process Map

A process map is a visual representation of how work moves through a process. It can show steps, decisions, handoffs, actors, inputs, outputs, delays, and the relationships among them.

Put the work on the page so the team can see it together.

See the complete working entry Review this term

Reviewed in POC

Prompt

A prompt is input used to guide what an AI system produces. Depending on the system, a prompt may arrive as text, spoken language, an image, a file, or a combination of media. It may contain a request, instruction, question, context, example, or constraint.

A prompt steers the work. It does not verify the result.

See the complete working entry Review this term

Reviewed in POC

Prompt Injection

Prompt injection is an attack in which instructions placed in user input or retrieved content attempt to override, redirect, or exploit an AI system's intended instructions and tool access.

Untrusted content tries to become trusted instruction.

See the complete working entry Review this term

Reviewed in POC

Proof of Concept

A proof of concept is a bounded experiment used to test whether an idea is feasible enough to justify further work, without claiming production readiness or validated value at scale.

A POC proves one question is worth the next question.

See the complete working entry Review this term

Reviewed in POC

Red Teaming

Red teaming is an authorized, structured exercise in which participants act like adversaries, harmful users, or failure-seeking evaluators to expose vulnerabilities, unsafe behavior, and weaknesses in a system or its safeguards.

Think like the threat, within rules.

See the complete working entry Review this term

Reviewed in POC

Responsible AI

Responsible AI is an umbrella commitment to develop and use AI in ways that are lawful, safe, fair, transparent, accountable, privacy-aware, and appropriate to the people and context affected.

Principles become responsibility only when they govern real choices.

See the complete working entry Review this term

Reviewed in POC

RLHF

RLHF stands for reinforcement learning from human feedback, a method that uses human preferences to help train a model toward more desirable responses.

Human preferences train a reward; the reward steers the model.

See the complete working entry Review this term

Reviewed in POC

Robot

A robot is a machine that can sense, compute, and act in the physical world, and the word is often used loosely in older conversations about artificial intelligence.

A robot senses, computes, and acts in the physical world.

See the complete working entry Review this term

Needs review

Robustness

Robustness is the degree to which a model or system continues to perform acceptably when inputs, conditions, users, or attacks differ from the ordinary cases used in development.

Robust means the system can take a hit without losing the task.

See the complete working entry Review this term

Reviewed in POC

Root-Cause Hypothesis

A root-cause hypothesis is a proposed explanation for what is actually producing an observed problem, stated in a form that can be tested. Until evidence supports it over the competing explanations, it stays a hypothesis, not a finding.

A cause is a hypothesis until evidence holds it up.

See the complete working entry Review this term

Reviewed in POC

Safe Curiosity

Safe curiosity is exploring a tool or idea through small, low-consequence experiments while protecting sensitive information, checking the result, and keeping important decisions human-owned.

Explore small. Protect the real. Check what happens.

See the complete working entry Review this term

Reviewed in POC

Sanitized Information

Sanitized information has been changed to remove or replace identifying, confidential, sensitive, or unnecessary details before it is shared. Sanitization can reduce risk, but it does not automatically make information anonymous, authorized, or safe.

Keep the pattern; remove the person.

See the complete working entry Review this term

Reviewed in POC

Scaling Laws

Scaling laws are empirical relationships that describe how selected measures of model performance or training behavior change as factors such as compute, data, and model size increase under specified conditions.

Scaling laws describe observed trends under stated conditions.

See the complete working entry Review this term

Reviewed in POC

Scenario

A scenario is a described situation, worked out in enough detail that people can reason about it — used for practice, planning or evaluation rather than as a record of something that happened. In the Learning Lab a scenario is always fictional, sanitized, or used with consent, so that no real client, colleague or customer is exposed by the exercise.

A safe story for practicing real judgment.

See the complete working entry Review this term

Reviewed in POC

Sensitive Information

Sensitive information is information that could create privacy, safety, financial, legal, employment, security, reputational, or relational harm if it is exposed, combined, or misused. What counts as sensitive depends on the people, setting, agreements, and consequences.

If exposure could cause harm, pause before sharing.

See the complete working entry Review this term

Reviewed in POC

Sensor

A sensor is a device that captures signals from the world, such as light, sound, motion, or temperature, so an AI system can perceive more than text alone.

A sensor turns a physical event into a usable signal.

See the complete working entry Review this term

Reviewed in POC

Service-Level Agreement

A service-level agreement is a contractual commitment defining specified service measures, targets, responsibilities, remedies, and methods of measurement between a provider and customer.

The SLA states which service promise is measurable and what happens if it is missed.

See the complete working entry Review this term

Reviewed in POC

Shadow AI

Shadow AI is the use of AI tools, accounts, integrations, or workflows outside an organization's approved governance, security, procurement, or visibility processes.

If the organization cannot see the AI use, it cannot govern the risk or learn from the need.

See the complete working entry Review this term

Reviewed in POC

State Transition Protocol

A state transition protocol is the agreed plan for moving a system from how it works now to how it will work next: the order of the steps, who owns each decision, how the new arrangement is verified as actually working, and how to get back if it doesn't. In most workplaces the same thing is called a change management plan, a cutover plan, or a runbook.

Move, verify, own, and know how to return.

See the complete working entry Review this term

Reviewed in POC

Stochastic Process

A stochastic process is a collection of random variables indexed by time, space, or another set. It models how an uncertain quantity or system can vary across that index under a probability law. It is also commonly called a random process.

A random variable is one uncertain snapshot. A stochastic process is an indexed series of uncertain snapshots together with the probability structure connecting them.

See the complete working entry Review this term

Reviewed in POC

Sycophancy

Sycophancy is AI behavior that excessively agrees with, flatters, or validates a user, including when evidence or sound judgment should lead the system to challenge or correct the user's view.

Helpful is not the same as telling the user what they want to hear.

See the complete working entry Review this term

Reviewed in POC

Synchronization Cascade

A synchronization cascade is a surge of same-direction behavior that happens when many independent actors read the same signal at nearly the same time and each reaches the same conclusion about an opportunity or a risk. Nobody coordinates; the timing does the work. Related ideas go by other names elsewhere — information cascades and herd behavior in economics, the thundering herd problem in computing.

Many actors see the same signal and move together.

See the complete working entry Review this term

Reviewed in POC

Synthetic Data

Synthetic data is artificially generated data designed to reproduce selected characteristics or situations for training, testing, simulation, analysis, or other uses.

Synthetic data is created to stand in for selected parts of reality.

See the complete working entry Review this term

Reviewed in POC

System

A system is a set of interacting components organized to achieve a purpose. In a work setting, those components can include people, tools, information, rules, processes, facilities, and surrounding conditions.

The system is the whole arrangement.

See the complete working entry Review this term

Reviewed in POC

System Prompt

A system prompt is a set of higher-priority instructions supplied by a model provider or application developer to shape an AI assistant's role, behavior, boundaries, and response style.

The application sets the role before the user makes the request.

See the complete working entry Review this term

Reviewed in POC

Token

A token is one of the small pieces an input is broken into before an AI model can process it. In text, a piece may be a whole word, part of a word, a punctuation mark, or a space — the model reads and produces these pieces rather than whole words. The program that does the breaking up is called a tokenizer.

Tokens are the pieces the model processes, not a fixed word count.

See the complete working entry Review this term

Reviewed in POC

Trust Borrowing

Trust borrowing is what happens when uncertain or machine-generated information gains credibility from the authoritative system, brand, or interface in which it appears, rather than from evidence of its own.

A trusted frame can lend trust to an untested claim.

See the complete working entry Review this term

Reviewed in POC

Vendor Lock-In

Vendor lock-in is a condition in which switching from a provider becomes costly, risky, slow, or impractical because important technology, data, skills, contracts, or workflows depend on that provider.

Adoption cost gets you in; switching cost determines how easily you get out.

See the complete working entry Review this term

Reviewed in POC

Verification

Verification is checking a claim or result against evidence that can actually settle it: a primary source, a system of record, direct observation, an approved standard, or a knowledgeable human reviewer. The higher the cost of being wrong, the more that evidence needs to be independent of whatever produced the claim — and at that point one knowledgeable person's say-so may not be enough on its own.

Match the check to the claim—and to the cost of being wrong.

See the complete working entry Review this term

Reviewed in POC

Workflow

A workflow is the operational path work follows through a system. It shows the sequence of steps, handoffs, decisions, people, tools, and information involved.

Follow the work to see what actually happens.

See the complete working entry Review this term

Reviewed in POC

Workslop

Workslop is AI-generated work that appears polished but lacks the substance needed to advance the task, leaving others to interpret, correct, or complete it.

Polished output can still contain unfinished thinking.

See the complete working entry Review this term

Reviewed in POC

World Model

A world model is a representation — learned from experience or built by hand — that a system uses to predict how an environment may change and how its own actions may affect future states.

The system acts on its model of the world, not the world directly.

See the complete working entry Review this term

Reviewed in POC

Zero-Shot Learning

Zero-shot means asking a model to do a task from instructions alone, with no examples. Few-shot means including a small number of worked examples in the same request. In neither case is the model retrained — the examples only shape the answer to that one request.

Zero-shot tells; few-shot shows.

See the complete working entry Review this term