AI Terminology Decoded
LLMs, tokens, fine-tuning, RAG, agents, hallucinations: AI conversations are full of terms most people nod along to. This course is a rapid, plain-language glossary that lets non-technical professionals hold their own in meetings and vendor calls. Each term comes with what it really means, why it matters, and the question it lets you ask. You'll finish able to talk about AI with confidence and spot vague claims.
Course details
Duration
57m 21s
Lectures
15
Video
1 hour
Quizzes
1
Level
Beginner
Course Description
You don't need to be an engineer to understand AI, but you do need the vocabulary. This course decodes the terms that come up in strategy meetings, product demos, and vendor contracts, and turns each one into a practical question you can ask.
What you'll learn
- How AI, machine learning, and generative AI relate, and how models are trained and used
- Large language models, tokens, context windows, temperature, and model versions
- How tools are built: fine-tuning, RAG, agents, APIs, copilots, and multimodal AI
- The risk vocabulary: hallucination, bias, drift, alignment, guardrails, and human-in-the-loop
- Data privacy terms like training on your data, retention, and zero-data-retention
- Five vendor questions you can use in any AI conversation
Who it's for
Managers, buyers, and non-technical staff who need to follow, question, and contribute to conversations about AI.
15 short video lessons in three modules: The Building Blocks, How the Tools Are Built, and The Risk Vocabulary.
Module 1: The Building Blocks
-
1.1 AI, Machine Learning, and Generative AIThe nesting-doll relationship between the three terms everyone mixes up — in one picture.
In this lesson:
- Define AI, machine learning, and generative AI
- Explain how the three nest inside each other
- Correct the most common mix-up in meetings
-
1.2 Models, Training, and Inference
What a model is, how it's built once, and what happens each time you use it.
In this lesson:
- Define a model in plain language
- Separate training from inference
- Explain why using a model doesn't teach it
-
1.3 Large Language Models and Tokens
Why the model reads in pieces, and what "context window" means for your long document.
In this lesson:
- Define a large language model
- Explain tokens and why they matter for cost and limits
- Use the context window concept to avoid a common failure
-
1.4 Prompts, Outputs, and Temperature
The dials you can and can't see — and why the same prompt gives different answers.
In this lesson:
- Define prompt, completion, and system instruction
- Explain temperature and why outputs vary
- Know which dials a user controls and which the product sets
-
1.5 Parameters, Versions, and "Frontier Models"
What the numbers mean, why bigger isn't always better for your task, and how to read a model name.
In this lesson:
- Explain parameters as a rough measure of model size
- Decode model names and version labels
- Match model size to the task instead of defaulting to the biggest
Module 2: How the Tools Are Built
-
2.1 Fine-Tuning and Custom Models
Teaching a model your style versus your facts — and why "custom model" rarely means what it sounds like.
In this lesson:
- Define fine-tuning and what it changes
- Distinguish teaching style from supplying facts
- Question a "custom model" claim precisely
-
2.2 RAG: Retrieval-Augmented Generation
How a tool "knows" your documents without being trained on them — and what to ask about it.
In this lesson:
- Explain retrieval-augmented generation in plain language
- Know why it's the standard way tools use company documents
- Ask the three questions that reveal a RAG system's quality
-
2.3 Agents and Automation
A model that takes actions instead of just answering — and why that changes the risk conversation.
In this lesson:
- Define an AI agent and how it differs from a chat assistant
- Explain why agents raise the stakes
- Ask the control questions before an agent is deployed
-
2.4 APIs, Integrations, and Copilots
The plumbing behind "AI-powered" features — and what it tells you about where your data goes.
In this lesson:
- Define API, integration, and copilot in plain terms
- Trace where data goes when a feature calls a model
- Ask the data-flow question about any AI feature
-
2.5 Multimodal: Text, Image, Audio, Video
One term, many capabilities, different risks — what multimodal means and the questions each mode raises.
In this lesson:
- Define multimodal and name the common modes
- Match each mode to its business use and its risk
- Recognize deepfake vocabulary as part of the same family
Module 3: The Risk Vocabulary
-
3.1 Hallucination
Confident invention — what the word means precisely, why it happens, and why it matters in a vendor meeting.
In this lesson:
- Define hallucination precisely
- Explain why it's a property of prediction, not a bug
- Use the term to ask the right vendor question
-
3.2 Bias, Drift, and Alignment
Three words that sound abstract until you know what each names — and the question each should trigger.
In this lesson:
- Define bias in AI output and where it comes from
- Define drift and why deployed systems need monitoring
- Define alignment and what it does and doesn't guarantee
-
3.3 Data Privacy Terms: Training, Retention, Zero-Data-Retention
The three phrases to find in any AI contract or settings page — and what each one actually commits to.
In this lesson:
- Define training on your data, retention, and zero-data-retention
- Find the three terms in a contract or settings page
- Ask the follow-up that closes each loophole
-
3.4 Guardrails, Red-Teaming, and Human-in-the-Loop
The safety vocabulary — what each term means, what it proves, and what it doesn't.
In this lesson:
- Define guardrails, red-teaming, and human-in-the-loop
- Know what each term does and doesn't prove
- Ask for the evidence behind each claim
-
3.5 Talking AI with Confidence
The glossary card, the five questions for any vendor, and a self-test — the finale.
In this lesson:
- Recall the course glossary from a one-card summary
- Use the five vendor questions in any AI conversation
- Complete AI Terminology Decoded
AI Terminology Decoded — Final Exam
Learn at your own pace
Study the materials at your convenience, you can always return to them to repeat
Certificate of completion
Complete every lesson and pass the final exam to earn a RiseCert certificate of completion you can download and share.
Your Name
AI Terminology Decoded
RISECERT
CERTIFICATE
For demonstrated knowledge and successful completion of all course requirements for:
MM.DD.YYYY
Certificate Date
Ready to get started?
Enroll in AI Terminology Decoded and earn your RiseCert certificate of completion.
Got Questions? We've Got Answers!
Answers to the most common questions about this course.
What is the AI Terminology Decoded course?
AI Terminology Decoded is a RiseCert online course that turns the twenty or so terms in nearly every AI conversation, from LLMs and tokens to RAG, agents, and hallucination, into plain language. Each term comes with the question it lets you ask. It includes 15 video lessons (just under an hour) and a final exam.
Who should take an AI terminology course?
Managers, buyers, team leads, and non-technical staff who sit in strategy meetings, product demos, or vendor calls where AI terms fly by. It suits anyone who needs to follow, question, and contribute to AI conversations, or evaluate an AI product, without becoming an engineer.
What's the difference between AI, machine learning, and generative AI?
They nest inside each other. AI is the broad field: software that does things we associate with human intelligence. Machine learning is the main method: learning patterns from data instead of hand-written rules. Generative AI is one branch of machine learning that creates new content, such as text, images, or code, rather than only classifying or predicting.
What is RAG (retrieval-augmented generation)?
RAG is how most tools answer questions about your company's documents. The system retrieves the most relevant passages, adds them to the prompt, and the model generates an answer grounded in them. The model isn't trained on your documents, so answers can stay current and cite their source. AI Terminology Decoded covers RAG in Module 2.
What is an AI hallucination?
A hallucination is content an AI model presents as fact, such as a citation, figure, quote, or feature, that isn't in the material it was given and isn't true. It happens because the model predicts likely language and truth isn't a step in that process. It can be reduced, not eliminated, so a claim of zero hallucinations is a red flag.
Which AI terms does the course cover?
Three modules. The Building Blocks: AI, machine learning, generative AI, models, training and inference, LLMs, tokens, context windows, temperature, and parameters. How the Tools Are Built: fine-tuning, RAG, agents, APIs, integrations, copilots, and multimodal. The Risk Vocabulary: hallucination, bias, drift, alignment, data retention, zero-data-retention, guardrails, red-teaming, and human-in-the-loop.
Do I need a technical background to take AI Terminology Decoded?
No. The course is written for non-technical professionals and explains every term in plain language with workplace examples like vendor pitches, contracts, and meetings. There's no math or coding. If you've used a chat assistant or sat through an AI demo, you already have all the background you need.
What will I be able to do after completing the course?
You'll be able to place any AI term in its family, read a model name, tell fine-tuning from retrieval, trace where your data goes through four stops, recognize weak human-in-the-loop claims, and find the training, retention, and zero-data-retention terms in a contract. You'll also leave with five questions to use in any vendor conversation.
How is this different from reading an AI glossary online?
A glossary gives definitions. AI Terminology Decoded pairs each term with why it matters and the specific question it lets you ask, such as whether a vendor retrains on your data or just retrieves it. It also teaches how the terms fit together, so you can spot vague or misused claims in a real pitch.
Does the course recommend specific AI tools or vendors?
No. AI Terminology Decoded is vendor-neutral and endorses no product or model. It deliberately teaches concepts rather than product names, because model names and version numbers change every few months while the concepts, such as the four slots in a model name, stay the same.
Is the course content up to date?
Yes. The course reflects AI tools and vendor practices as of 2026. Because product and model names change so quickly, it focuses on durable concepts and questions, such as training versus inference, retention terms, and hallucination rates, that stay useful as specific tools and versions come and go.
How long does the course take to complete?
AI Terminology Decoded has just under an hour of video across 15 lessons, plus a final exam. Lessons run about four minutes each and are organized into three modules: The Building Blocks, How the Tools Are Built, and The Risk Vocabulary. Most learners finish in one or two sittings.
Is the course online and self-paced?
Yes. The course is 100% online and self-paced, and your access does not expire. You can start, pause, and resume at any time on a computer, tablet, or phone with a modern web browser. There's nothing to install.
Is there a final exam?
Yes. The course ends with a 30-question multiple-choice final exam covering all three modules. You need a score of 80% or higher (24 of 30) to pass, and you can retake the exam as many times as you need. Finish all 15 lessons and pass the exam to complete the course.
Do I get a certificate of completion?
Yes. After you complete all 15 lessons and pass the final exam, you can download a RiseCert certificate of completion from your account. It is a course completion certificate, not a license, a professional certification, or a continuing education (CE) certificate.
Does AI Terminology Decoded count for continuing education (CE or CPE) credit?
No. The course is not currently approved or accredited for continuing education credit, including CE, CPE, or CEU hours, by any state, licensing body, or professional association. If you need credit for a license or certification, check with your credentialing body for approved courses.
Is the course technical, legal, or procurement advice?
No. AI Terminology Decoded is educational and is not legal, technical, security, or procurement advice. It teaches the vocabulary and questions for evaluating AI tools, but decisions about specific contracts, data handling, or compliance should involve your legal, security, and IT teams.
How much does the course cost, and can I enroll my team?
The course costs $99 per learner as a one-time purchase, with no subscription or recurring fees. To enroll a team or department, contact RiseCert through the Contact Us page at risecert.org.
What is the refund policy?
All sales are final and no refunds are issued, as stated in the RiseCert Terms of Use. Please review the course description, lesson list, and these FAQs before purchasing, and contact RiseCert with any questions first.
What questions should I ask an AI vendor?
AI Terminology Decoded teaches five: what predicts and what generates; whose model it is and how it's customized; where your data goes and how long it's kept; the hallucination rate on tasks like yours and what happens when the answer isn't there; and exactly where a human is in the loop, what they see, and whether they can stop it.