Practical Applications of Prompt (PXMO) - Pre-Assessment Review

A comprehensive vocabulary review of prompt engineering components, prompting patterns, AI biases, ethical principles, and multimodal technologies from the PXMO Pre-Assessment.

Persona

A prompt engineering component that assigns a specific role or perspective (such as "You are a lawyer") to direct the model to generate highly relevant and appropriate responses.

Instructions

The component of a prompt that directs the artificial intelligence to perform a specific task or action (such as "Find bike paths near Minneapolis").

Constraints

Rules and boundaries defined within a prompt—such as character counts, length restrictions, or prohibited language—that establish the scope and prevent overly broad or nonsensical output.

Delimiters

Punctuation marks, symbols, or formatting tags inserted between distinct sections of a prompt to ensure the model processes instructions separately from background context.

Tree of Thought (TOT)

A prompting technique that instructs an AI system to generate and explore multiple possible paths, endings, or solutions before selecting the most suitable response.

Generated Knowledge Prompting

A prompting technique that elicits background information or facts from an initial prompt and then uses that knowledge to guide the response to a follow-up prompt.

Self-Consistency

A prompting technique that involves running the exact same prompt multiple times and taking the most recurring or consistent result as the final answer.

Task Decomposition (Multi-Step Prompting)

The strategy of breaking a complex objective into sequential prompts—such as drafting content first for accuracy and rewriting in a second step for clarity and tone.

Generative AI Processing Mechanism

The computational method by which algorithms interpret prompt input and identify patterns across training data to synthesize new text, code, or images.

Reinforcement from Human Feedback

A process where human evaluators rate model outputs as helpful or unhelpful, providing labeled feedback that acts as rewards and penalties to refine the system toward better clarity and tone.

Algorithmic Bias

Systematic discrimination or skewed results arising from patterns in historical training data, such as disproportionately advising loan rejections for residents of a particular geographic area.

Measurement Bias

A systematic error where an AI model consistently produces measurements or forecasts that deviate by a regular margin (such as consistently underpredicting or overpredicting temperatures by five degrees).

Confirmation Bias (in AI Prompting)

The user behavior of repeatedly querying an AI system until it outputs an answer that aligns with the user's pre-existing belief or favored outcome.

Transparency

The ethical principle requiring organizations to openly disclose AI methodologies, reveal whether content or evaluations were AI-generated, and inform users if they are interacting with an AI system rather than a human.

Accountability

The ethical principle focused on establishing clear governance mechanisms to assign legal and organizational responsibility for the actions, outcomes, and potential harms of an AI system.

Privacy

The ethical principle of implementing safeguards and security measures to prevent the unauthorized access, disclosure, or misuse of sensitive personal or organizational data in AI systems.

Fairness

The ethical AI principle ensuring that systems treat individuals equitably without discriminating against members of federally protected classes or specific demographic groups.

Optical Character Recognition (OCR)

An AI technology that analyzes printed, typed, or handwritten text on physical documents—such as receipts, shipping labels, and legal contracts—and converts it into searchable digital data.

Speech Recognition

An AI tool that converts spoken verbal commentary into written text or structured data entries in real time.

Voice Recognition

An advanced AI capability that analyzes biometric vocal characteristics to identify and verify the name or identity of the person speaking.

Speech Synthesis

An advanced AI application that translates digital written text into audible spoken language.

Image Prompt Specificity

The practice of clearly defining the main subject, setting, and environment in a text prompt to provide precise guidance for AI image generators.