Glossary¶

This glossary contains key terms and definitions used throughout the SEIS 666: Digital Transformation 2.0 with Generative AI course. Definitions follow ISO 11179 metadata registry standards: precise, concise, distinct, non-circular, and free of business rules.

A¶

Accountability¶

The obligation to explain AI system decisions and accept responsibility for their outcomes.

Adversarial Testing¶

A security evaluation method that simulates attacks to identify vulnerabilities in AI systems.

AI Agents¶

Software programs that autonomously perform tasks by perceiving their environment and taking actions to achieve goals.

AI Bias¶

Systematic errors in AI outputs that reflect prejudices in training data or algorithm design.

AI Champions¶

Designated individuals who advocate for AI adoption and support implementation within their organizations.

AI Ethics¶

The branch of ethics examining moral issues arising from the development and deployment of artificial intelligence.

AI Governance¶

The framework of policies, processes, and organizational structures that guide responsible AI development and use.

AI Infrastructure¶

The computing resources, platforms, and tools required to develop, deploy, and maintain AI systems.

AI Maturity Model¶

A framework for assessing an organization's capability to effectively implement and scale AI initiatives.

AI Policy¶

Formal guidelines that define acceptable use, development standards, and compliance requirements for AI systems.

AI Regulations¶

Legal requirements governing the development, deployment, and use of artificial intelligence systems.

AI Roadmap¶

A strategic plan outlining the timeline, milestones, and resources for implementing AI initiatives.

AI Strategy¶

A comprehensive plan aligning AI investments and capabilities with organizational goals and competitive positioning.

AI Transformation¶

The fundamental restructuring of business operations and models through the integration of AI technologies.

AI Use Case¶

A specific business scenario where AI can be applied to solve a problem or create value.

AI-Augmented Workforce¶

Human workers whose capabilities are enhanced through collaboration with AI tools and systems.

Anthropic¶

An AI safety company that develops Claude and other large language models focused on being helpful, harmless, and honest.

Anthropic API¶

The programming interface for integrating Claude models into applications and workflows.

API Authentication¶

The process of verifying the identity of applications or users accessing an API.

API Endpoints¶

Specific URLs where API requests are sent to access particular functions or resources.

API Fundamentals¶

The basic concepts and principles underlying application programming interfaces.

API Keys¶

Unique identifiers used to authenticate and authorize access to API services.

API Pricing¶

The cost structure for using API services, typically based on usage metrics like token count.

Artificial Intelligence¶

Computer systems designed to perform tasks that typically require human intelligence.

Attention Mechanism¶

A neural network component that allows models to focus on relevant parts of input when generating output.

Audio AI¶

Artificial intelligence systems specialized in processing, generating, or analyzing sound and speech.

Augmented Intelligence¶

AI systems designed to enhance human decision-making rather than replace human judgment.

Autonomous Systems¶

Technology systems capable of operating independently without continuous human intervention.

B¶

Best Practices¶

Proven methods and techniques that consistently produce superior results in a given domain.

Bias Detection¶

The process of identifying systematic errors or unfairness in AI model outputs.

Bias Mitigation¶

Techniques and processes used to reduce or eliminate bias in AI systems.

Blockchain and AI¶

The integration of distributed ledger technology with artificial intelligence applications.

Business Drivers¶

Factors that motivate organizations to pursue specific initiatives or strategies.

Business Model Innovation¶

The creation of new ways to create, deliver, and capture value through AI-enabled capabilities.

C¶

Capstone Project¶

A comprehensive final project integrating multiple course concepts to demonstrate mastery.

Case Study Analysis¶

The systematic examination of real-world examples to extract insights and lessons.

Chain-of-Thought¶

A prompting technique that encourages models to show step-by-step reasoning before providing answers.

Change Management¶

The structured approach to transitioning individuals and organizations to desired future states.

ChatGPT¶

OpenAI's conversational AI assistant based on GPT models, designed for interactive dialogue.

Claude¶

Anthropic's family of large language models designed to be helpful, harmless, and honest.

Claude 3 Opus¶

The most capable model in the Claude 3 family, optimized for complex tasks requiring deep analysis.

Claude 3 Sonnet¶

A balanced model in the Claude 3 family offering strong performance with efficient processing.

Cloud AI Services¶

AI capabilities delivered through cloud computing platforms on a subscription or usage basis.

Competitive Advantage¶

A capability or asset that enables an organization to outperform its competitors.

Content Moderation¶

The process of monitoring and filtering AI-generated content to ensure appropriateness.

Context Window¶

The maximum amount of text a language model can process in a single interaction.

Converging Technologies¶

The integration of multiple technology domains to create new capabilities and applications.

Copyright AI Content¶

Legal protections and considerations for content created by or with assistance from AI.

Cosine Similarity¶

A mathematical measure of similarity between two vectors based on the cosine of their angle.

Cost Optimization¶

Strategies to minimize expenses while maintaining desired performance in AI operations.

Creativity Enhancement¶

The use of AI tools to augment human creative capabilities and output.

Custom GPT¶

A specialized version of ChatGPT configured for specific use cases or domains.

Customer Experience AI¶

AI applications designed to improve customer interactions and satisfaction.

D¶

DALL-E¶

OpenAI's AI system that generates images from text descriptions.

Data Privacy¶

The protection of personal information from unauthorized access or disclosure.

Data Security¶

Measures taken to protect data from unauthorized access, corruption, or theft.

Deep Learning¶

A subset of machine learning using neural networks with multiple layers to learn complex patterns.

Diffusion Models¶

Generative AI models that create images by iteratively removing noise from random patterns.

Digital Capability Model¶

A framework for assessing and developing organizational digital competencies.

Digital Economy¶

An economy based on digital computing technologies and internet-enabled commerce.

Digital Maturity¶

The degree to which an organization has developed capabilities to leverage digital technologies effectively.

Digital Transformation¶

The integration of digital technology into all areas of business operations and value delivery.

Digitalization¶

The use of digital technologies to change business models and create new value opportunities.

Digitization¶

The conversion of analog information into digital format.

E¶

Edge AI¶

Artificial intelligence processing performed locally on devices rather than in centralized cloud servers.

Embeddings¶

Dense vector representations of data that capture semantic meaning for machine learning.

Enterprise AI¶

Large-scale AI implementations across organizational functions and processes.

EU AI Act¶

European Union legislation establishing rules for the development and use of AI systems.

Executive Sponsorship¶

Active support from senior leadership for initiatives, providing resources and organizational alignment.

Explainability¶

The ability to describe AI decision-making processes in terms humans can understand.

F¶

Factual Accuracy¶

The degree to which AI-generated content correctly represents verifiable information.

Failure Patterns¶

Common characteristics or sequences of events that lead to unsuccessful outcomes.

Feasibility Analysis¶

Assessment of whether a proposed initiative can be successfully implemented.

Few-Shot Prompting¶

A technique providing a small number of examples to guide model responses.

Finance AI¶

Artificial intelligence applications in financial services and investment management.

Fine-Tuning¶

The process of adapting a pre-trained model to specific tasks or domains with additional training.

Frozen in Time¶

The characteristic of AI language models having a knowledge cutoff date, meaning they lack awareness of events occurring after their training data was collected.

Future of Work¶

The evolution of employment, skills, and workplace dynamics driven by technological change.

G¶

GAI Center of Excellence¶

An organizational unit dedicated to developing and scaling generative AI capabilities.

GAICoE Charter¶

A formal document defining the mission, scope, and governance of an AI Center of Excellence.

Adherence to the European Union's General Data Protection Regulation requirements.

Gemini Pro¶

Google's capable multimodal AI model for a range of tasks.

Gemini Ultra¶

Google's most advanced multimodal AI model for complex reasoning and generation.

Generative AI¶

AI systems capable of creating new content such as text, images, audio, or code.

Google Gemini¶

Google's family of multimodal AI models capable of processing text, images, and other data types.

GPT Actions¶

Custom functions that extend GPT capabilities to interact with external services.

GPT Builder¶

OpenAI's interface for creating and configuring custom GPT applications.

GPT-4¶

OpenAI's large multimodal model capable of processing text and images.

GPT-4 Turbo¶

An optimized version of GPT-4 with improved speed, longer context, and lower cost.

GPT-4 Vision¶

GPT-4's capability to analyze and respond to image inputs.

GPT-4o¶

OpenAI's omni model with native multimodal capabilities across text, vision, and audio.

Grounding¶

Techniques to connect AI outputs to verified factual information sources.

H¶

Hallucination¶

AI model outputs that appear plausible but contain fabricated or incorrect information.

Healthcare AI¶

Artificial intelligence applications in medical diagnosis, treatment, and healthcare delivery.

Human-AI Collaboration¶

Work arrangements where humans and AI systems cooperate to achieve outcomes.

Hybrid AI¶

Systems combining cloud-based and edge-based AI processing capabilities.

I¶

Image Analysis¶

The automated extraction of information and insights from visual content.

Image Generation¶

The creation of new images by AI systems based on text prompts or other inputs.

Impact Assessment¶

Evaluation of the potential effects of an initiative on stakeholders and operations.

In-Context Learning¶

A model's ability to perform new tasks based on examples provided in the prompt.

Industry Use Cases¶

Sector-specific applications of AI that address common business challenges.

Inference¶

The process of using a trained model to generate predictions or outputs from new inputs.

Intellectual Property¶

Legal rights protecting creations of the mind, including AI-generated works.

IoT and AI¶

The integration of Internet of Things sensors and devices with artificial intelligence analytics.

J¶

Job Creation¶

The emergence of new employment opportunities resulting from AI adoption.

Job Displacement¶

The elimination of existing job roles due to AI automation.

JSON Output¶

Structured data format commonly used for API responses and data interchange.

K¶

Knowledge Bases¶

Organized collections of information used by AI systems for retrieval and reference.

L¶

Large Language Models¶

Neural networks trained on massive text datasets to understand and generate human language.

Latency¶

The time delay between sending a request and receiving a response from an AI system.

Lessons Learned¶

Insights gained from experience that can improve future performance.

Low-Code Platforms¶

Development environments enabling application creation with minimal programming.

M¶

Machine Learning¶

A subset of AI where systems improve performance through experience without explicit programming.

Manufacturing AI¶

Artificial intelligence applications in production, quality control, and supply chain management.

Markdown Output¶

Text formatted using Markdown syntax for structured document generation.

Max Tokens Parameter¶

An API setting that limits the length of generated responses.

Meta Llama¶

Meta's family of open-source large language models.

Midjourney¶

An AI image generation service known for artistic and creative outputs.

Mistral AI¶

A French AI company developing efficient open-source language models.

Mixtral¶

Mistral AI's mixture-of-experts model architecture for efficient inference.

Model Parameters¶

The learned weights and biases that define a neural network's behavior.

Multi-Head Attention¶

An attention mechanism that processes information through multiple parallel attention operations.

Multimodal AI¶

AI systems capable of processing and generating multiple types of data such as text, images, and audio.

Multimodal Applications¶

Software using AI to work with multiple data types simultaneously.

N¶

Neural Networks¶

Computing systems inspired by biological neural networks, composed of interconnected nodes.

No-Code AI Tools¶

AI applications that enable users to build solutions without writing code.

O¶

Open-Source Models¶

AI models with publicly available weights and code for community use and modification.

OpenAI¶

An AI research company that develops GPT models and ChatGPT.

OpenAI API¶

The programming interface for accessing OpenAI's language and multimodal models.

Operational Excellence¶

The consistent execution of business processes to deliver superior performance.

Organizational Change¶

The transformation of an organization's structure, culture, or processes.

Organizational Readiness¶

An organization's preparedness to successfully implement new initiatives.

Output Formatting¶

Techniques for structuring AI responses in specific formats.

P¶

Persona Design¶

The creation of defined AI personalities and communication styles through prompting.

Perplexity AI¶

An AI-powered search engine that provides answers with cited sources.

PII Protection¶

Safeguards for personally identifiable information in AI systems.

Pre-Training¶

The initial training phase where models learn general patterns from large datasets.

Prioritization Framework¶

A structured method for ranking initiatives based on defined criteria.

Productivity Enhancement¶

Improvements in output efficiency achieved through AI-assisted workflows.

Prompt Engineering¶

The practice of designing effective inputs to optimize AI model outputs.

Prompt Iteration¶

The process of refining prompts through successive improvements.

Prompt Libraries¶

Collections of tested prompts organized for reuse across applications.

Prompt Optimization¶

Techniques for improving prompt effectiveness and efficiency.

Prompt Templates¶

Reusable prompt structures with variable placeholders for customization.

Proprietary Models¶

AI models with restricted access controlled by their developers.

Q¶

Quick Wins¶

High-impact initiatives that can be implemented rapidly with minimal resources.

R¶

RAG¶

Retrieval-Augmented Generation: a technique combining information retrieval with text generation.

Rate Limiting¶

Controls that restrict the frequency of API requests to manage system load.

Red-Teaming¶

A security practice where teams attempt to find vulnerabilities in AI systems.

Reskilling¶

Training workers in new skills to adapt to changing job requirements.

Responsible AI¶

Development and deployment practices that ensure AI systems are ethical and beneficial.

REST API¶

An architectural style for web services using standard HTTP methods.

Retail AI¶

Artificial intelligence applications in retail operations and customer experience.

Retrieval Systems¶

Components that find and return relevant information from knowledge bases.

RLHF¶

Reinforcement Learning from Human Feedback: a training method using human preferences.

ROI Estimation¶

The calculation of expected return on investment for proposed initiatives.

Role Evolution¶

Changes in job responsibilities and requirements driven by technology adoption.

S¶

Safety Guardrails¶

Constraints built into AI systems to prevent harmful outputs.

Scaling AI¶

The process of expanding AI capabilities across an organization.

SDK¶

Software Development Kit: tools and libraries for building applications with APIs.

Search-Augmented Generation¶

AI text generation enhanced with real-time search results.

Self-Attention¶

An attention mechanism where a sequence attends to itself to capture relationships.

Self-Consistency¶

A prompting technique that generates multiple reasoning paths and selects the most common answer.

Semantic Search¶

Search technology that understands meaning and intent rather than just keywords.

Similarity Search¶

Finding items in a database based on their similarity to a query vector.

Skill Transformation¶

Changes in the skills required for jobs due to technological advancement.

Sora¶

OpenAI's text-to-video generation model.

Speech-to-Text¶

AI technology that converts spoken language into written text.

Stable Diffusion¶

An open-source AI model for generating images from text descriptions.

Stakeholder Engagement¶

The process of involving affected parties in planning and decision-making.

Stop Sequences¶

Character sequences that signal the model to stop generating output.

Strategic Initiatives¶

Long-term projects aligned with organizational strategy and goals.

Streaming Responses¶

API functionality that delivers outputs incrementally as they are generated.

Structured Output¶

AI responses formatted in predefined data structures.

Success Factors¶

Elements that contribute to achieving desired outcomes.

System Prompt¶

Instructions that define AI behavior and context for a conversation.

T¶

Temperature Parameter¶

An API setting that controls randomness in model outputs.

Text-to-Image¶

AI technology that generates images from textual descriptions.

Text-to-Speech¶

AI technology that converts written text into spoken audio.

Text-to-Video¶

AI technology that generates video content from text descriptions.

Throughput¶

The rate at which an AI system processes requests or generates outputs.

Token¶

The basic unit of text processed by language models, typically a word or subword.

Token Counting¶

Measuring the number of tokens in text for API usage and cost calculation.

Tokenization¶

The process of converting text into tokens for model processing.

Top-P Parameter¶

An API setting that controls output diversity by limiting cumulative probability.

Transformer Architecture¶

A neural network design using attention mechanisms for sequence processing.

Transparency¶

The quality of AI systems being open about their operations and limitations.

Tree-of-Thought¶

A prompting technique that explores multiple reasoning branches before selecting answers.

U¶

Upskilling¶

Training workers to enhance existing skills for improved performance.

Use Case Identification¶

The process of discovering opportunities for AI application in business contexts.

User Prompt¶

The input message provided by users to AI systems.

V¶

Value Creation¶

The process of generating benefits for customers and stakeholders.

Value Mapping¶

Connecting AI capabilities to specific business value and outcomes.

Vector Database¶

A specialized database optimized for storing and searching vector embeddings.

Vision Capabilities¶

AI functionality for processing and understanding image and video content.

Voice Cloning¶

AI technology that replicates a specific person's voice characteristics.

W¶

Workflow Automation¶

The use of technology to execute business processes with minimal human intervention.

X¶

xAI Grok¶

xAI's large language model designed for real-time information access.

Z¶

Zero-Shot Prompting¶

A technique where models perform tasks without task-specific examples.

This glossary contains 200 terms aligned with the SEIS 666 learning graph concepts.