// A–Z
Generative AI Terms
A
- AI Agent
- An AI system that can plan and carry out multi-step tasks, such as researching, filling in forms, or moving data between apps, with limited human input. Responsible use requires clear human approval points.
- AI Beta Team
- A small cross-functional group inside an organization that tests generative AI tools safely, documents best practices, and guides wider rollout. See the Beta Team Integration Program.
- AI Usage Policy
- An organization's internal rules for how generative AI may be used, covering approved tools, confidential data, human review, hallucinations, bias, and compliance.
- Answer Engine Optimization (AEO)
- Optimizing content so AI-powered search and chat tools select it as the direct answer to a question. Closely related to Generative Engine Optimization.
- Anthropic
- The AI company that develops the Claude family of models.
- Approved Tools
- The AI systems an organization has authorized employees to use, usually listed in its AI usage policy.
- Artificial Intelligence (AI)
- Computer systems designed to perform tasks that usually require human intelligence, such as reasoning, learning, and problem-solving.
B
- Bias
- Systematic errors in training data or model design that cause an AI system to produce unfair or skewed results.
C
- ChatGPhD
- YEN's term for an ongoing commitment to mastering AI in a business context, and a proficiency framework, modeled on martial-arts belts, for measuring practical AI skill.
- ChatGPT
- OpenAI's AI assistant, built on its GPT large language models, that generates conversational text and can analyze files, images, and data.
- Claude
- A family of conversational AI models developed by Anthropic, used for writing, analysis, coding, and research.
- Closed-Loop AI System
- An AI setup that runs in a secure, controlled environment without exposing data externally. Common in law firms, healthcare, and other confidentiality-sensitive settings.
- Confidential Data
- Sensitive company or client information that must be protected and should not be entered into unapproved AI tools.
- Confidentiality Risk
- The danger of exposing private or client data by entering it into public or improperly configured AI tools.
- Custom GPT (MyGPT)
- A customized version of ChatGPT set up with specific instructions, documents, or workflows for personal or organizational use.
D
- Data Privacy
- The protection of personal and sensitive information from unauthorized access or use.
- Deep Learning
- A branch of machine learning that uses multi-layered neural networks to model complex patterns in data.
- Deep Research
- An AI mode, offered in tools like ChatGPT, Gemini, and Claude, that runs a multi-step web research process and returns a cited report. Cross-checking results across models helps catch hallucinations.
- DeepSweep
- YEN's AI visibility measurement process. It asks the same customer questions many times across ChatGPT, Claude, and Gemini and reports Mention Rate, Top-3 Rate, rank, and share of voice. See GEO services.
- Diffusion Model
- A generative AI model, used by many image tools, that creates images by gradually transforming random noise into a picture.
E
- Embedding
- A way of representing words, sentences, or images as numerical vectors so AI models can measure how similar they are.
- Ethical AI
- The practice of building and using AI responsibly, addressing bias, privacy, transparency, and safety.
F
- Few-Shot Learning
- Giving a model a handful of examples in the prompt to guide the format or style of its response.
- Fine-Tuning
- Further training a pre-trained model on a smaller, specialized dataset so it performs better on a specific task.
- Fractional AI Leadership
- Part-time executive leadership, similar to a fractional COO, focused on bringing AI into an organization safely and strategically.
G
- GAN (Generative Adversarial Network)
- A deep learning model in which two neural networks, a generator and a discriminator, compete to produce realistic content.
- Gemini
- Google's family of multimodal AI models and the assistant built on them, formerly known as Bard.
- Generative AI
- A type of artificial intelligence that creates new content such as text, images, audio, video, or code. See What is Generative AI?
- Generative AI Accelerator
- YEN's structured training format that moves people and teams from basic awareness to hands-on AI use with accountability. See the Generative AI 101 Accelerator.
- Generative AI Mastermind
- A peer group in which business owners and professionals learn, share, and apply AI use cases together. See the AI Advanced Mastermind.
- Generative Engine Friendly
- YEN's term for a website built to be easily understood and cited by large language models like ChatGPT, Claude, and Gemini, not just by traditional search engines.
- Generative Engine Optimization (GEO)
- The practice of structuring a website's content, business facts, and structured data so AI answer engines can understand, trust, and cite it. See GEO services.
- Grok
- The AI assistant from Elon Musk's company xAI, integrated with the X platform.
H
- Hallucination
- When a generative AI model confidently produces false or fabricated information. Mitigation includes grounding answers in trusted sources, asking for citations, comparing models, and human fact-checking.
I
- IDEA Framework
- YEN's method for matching AI models to project stages, Ideate, Draft, Engineer, Assess, so premium reasoning models are reserved for the steps that need them.
L
- Large Language Model (LLM)
- A deep learning model trained on massive amounts of text to understand and generate human-like language. ChatGPT, Claude, and Gemini run on LLMs.
M
- Machine Learning (ML)
- A subset of AI in which systems learn patterns from data to make predictions or decisions without being explicitly programmed.
- Mention Rate
- The share of collected AI answers that name a business at least once. The baseline metric in YEN's DeepSweep reports.
- Meta AI (Llama)
- Meta's AI assistant and its Llama family of openly available language models.
- Microsoft Copilot
- Microsoft's AI assistant, integrated into Microsoft 365 apps such as Outlook, Word, Excel, and PowerPoint.
- Model Parameters (Weights)
- The internal values of a neural network that are adjusted during training as the model learns patterns.
- Multimodal AI
- AI that works across multiple types of data, such as text, images, audio, and video.
N
- Neural Network
- Algorithms loosely inspired by the human brain that process data through layers of interconnected nodes to recognize patterns.
O
- OpenAI
- The company that created ChatGPT, the GPT models, and other generative AI tools.
P
- People First, People Last
- YEN's core philosophy for responsible AI use: a person frames the request at the start and a person reviews the output at the end, so AI strengthens human decision-making instead of replacing it.
- PII (Personally Identifiable Information)
- Data that can identify an individual, such as names, Social Security numbers, or addresses.
- Prompt
- The text or instruction given to a generative AI model to guide its output.
- Prompt Engineering
- The practice of structuring prompts with context, a role, a clear task, a format, and a tone to improve the quality of AI results.
- Prompt Genius
- A YEN chatbot that helps users write stronger prompts.
- Prompt Injection
- Hidden instructions placed in a document, web page, or email that trick an AI tool into doing something the user did not intend.
- Proprietary Information
- Company-owned intellectual property or trade secrets.
R
- RAG (Retrieval-Augmented Generation)
- A technique that connects a language model to an external knowledge source, often a vector database, so it retrieves relevant facts before generating an answer.
S
- Shadow AI
- Employees' use of AI tools their organization has not approved.
T
- Token
- The basic unit of text, a word or part of a word, that language models read and generate. Usage limits and pricing are often measured in tokens.
- Training Data
- The text, images, or other inputs used to teach a machine learning model.
- Transformer Model
- The neural network architecture behind modern LLMs, designed to process sequences of text efficiently.
V
- Vector Database
- A database that stores embeddings so AI applications can run fast similarity searches.
- Voice-to-Voice AI
- Multimodal AI that both listens and speaks, used for assistants and client communication.
Z
- Zero-Shot Learning
- When a model completes a task without any examples, relying only on its general training.
// Learn by Doing
Turn Terms Into Skills
Knowing the vocabulary is step one. Our classes and team training put these concepts to work on real tasks.