AI & Prompt Engineering English Exercises — English for IT Professionals — English for IT

Intermediate

Writing Clear AI Prompts

Practice writing precise, unambiguous instructions for AI systems. Learn how role, context, format, and constraints affect output quality.

4 exercises
Intermediate

Evaluating LLM Outputs

Describe and critically assess AI-generated text. Use vocabulary like hallucination, grounding, coherence, and relevance.

3 exercises
Beginner

AI & ML Vocabulary

LLM, RAG, fine-tuning, embeddings, tokens, context window, temperature — the terms every modern developer needs.

4 exercises
Advanced

Discussing AI Risks & Ethics

Articulate concerns about bias, hallucinations, data privacy, and responsible AI use in professional English.

2 exercises
Advanced

AI System Design Language

Describe RAG architectures, vector databases, inference pipelines, and model evaluation strategies clearly.

2 exercises

Useful language for AI discussions

Describing prompts

"The prompt specifies the role and constrains the output format."
"Zero-shot means no examples are provided in the prompt."
"I used chain-of-thought prompting to improve reasoning."
"The system prompt sets the assistant's persona."

Evaluating outputs

"The model hallucinated several facts here."
"The output lacks grounding — there's no source for this claim."
"The response is coherent but not relevant to the question."
"Increasing the temperature makes outputs more creative but less reliable."

Discussing AI systems

"We use RAG — retrieval-augmented generation — to ground responses in our docs."
"The context window limits how much we can pass to the model."
"We're fine-tuning on domain-specific data to improve accuracy."
"Embeddings are vector representations of text for semantic search."