DeepDojo.AI
CoursesPricingGuidesBlog

Open courseSkill DevelopmentAI Courses

Basic Prompting Techniques

Write clear requests, ground answers in evidence, and check complex work one stage at a time

Start learning
All chapters
4
Locale
English
Course type
Open course

Learn to address four common problems when using AI: vague requests, unsupported answers, poorly chosen reasoning methods, and tasks too large to complete in one attempt. Practice rewriting prompts, setting source requirements, choosing a reasoning method, and dividing complex work into stages you can check.

Best for

  • Beginners who want to use AI more effectively and get better answers
  • People who use AI for studying, writing, office work, or organizing information
  • People who want fewer off-topic, fabricated, or vague answers

Problems solved

  • Vague requests leave AI guessing and often lead to off-topic answers
  • You need accurate facts but have not specified sources or limits on inference
  • You ask for an immediate answer when the task needs careful reasoning or judgment
  • Tackling complex work in one attempt makes omissions and errors hard to find

Outcomes

  • Specify the task, context, constraints, examples, and output you need; remove redundant or conflicting instructions
  • Specify sources, supporting quotations, and rules for missing information; distinguish facts, summaries, and inferences
  • Decide when to reason step by step, start with principles, solve independently, or check an existing answer
  • Divide complex work into stages, each with an intermediate result, review criteria, and conditions for continuing

Highlights

  • Four lessons address clear requests, evidence, reasoning, and task breakdown
  • Examples cover study, writing, and organizing material for everyday AI users
  • Find gaps through flashcards, prompt revisions, questions about your reasoning, and diagnostic feedback

All chapters

4 chapters

Lesson 1: Make Your Request Clear

Use the prompt-review framework from the tutorial to identify what is unclear in a request, then rewrite it with a clearer task, context, constraints, source boundaries, and output structure.

Lesson 2: Ground Answers in Evidence

Use the tutorial's hallucination-prevention checks to decide whether a task needs reference text, then write a prompt with clear source, refusal, quotation, and inference boundaries.

Lesson 3: Let the Model Think Before It Answers

Use the tutorial's selection framework to decide when a task needs deliberate reasoning and choose an appropriate method, such as step-by-step analysis, step-back prompting, independent solving, verification, path comparison, or tool use.

Lesson 4: Break Down Complex Tasks

Use the tutorial's decomposition framework to divide a complex task into checkable stages and explain each stage's goal, intermediate output, and condition for continuing.