Phase 2: Thinking & Reasoning Techniques · 30 min · ChatGPT, Claude, or Gemini (any)
Chain-of-Thought: Show Your Work
Asking AI to think out loud used to be the biggest trick in the book. Now it depends on which model you're using.
Hiring signal: What changes when you use this: you know when 'think step by step' helps and when it's redundant or counterproductive — which means you stop wasting tokens on a technique that modern models don't need.
What you will learn
- Understand what chain-of-thought means in plain English
- Know when 'think step by step' improves output and when it doesn't
- Understand why reasoning models make chain-of-thought redundant or counterproductive
- Apply chain-of-thought selectively based on the model and task
The Problem
For years, the most reliable way to get better answers from AI was to add four words: "think step by step." This technique — called chain-of-thought — asked the AI to show its reasoning before giving its answer, and it dramatically improved accuracy on complex questions.
But the landscape has changed. Modern reasoning models (OpenAI's o-series, Claude with extended thinking, Gemini's reasoning modes) now do this internally. Asking them to "think step by step" can actually make output worse — it's like telling a chef to explain how to boil water before they cook.
This lesson teaches you when chain-of-thought helps, when it's redundant, and when it hurts.
What Chain-of-Thought Means
Chain-of-thought is simple: instead of asking the AI for an answer, you ask it to reason through the problem first, then give the answer.
Standard prompt:
If I have a budget of $5,000 and I need to buy 3 laptops at $1,200 each, 2 monitors at $350 each, and 5 software licenses at $200 each, do I have enough left over for a $500 desk?
Chain-of-thought prompt:
If I have a budget of $5,000 and I need to buy 3 laptops at $1,200 each, 2 monitors at $350 each, and 5 software licenses at $200 each, do I have enough left over for a $500 desk? Think through this step by step before giving your answer.
The AI will calculate each line item, add them up, subtract from the budget, and then answer. By showing its work, it's less likely to make a calculation error — because each step is visible and the AI can self-correct.
Unlock the full lesson
You've read the first 2 sections. The rest of this lesson covers When It Helps, When It's Redundant or Counterproductive, The Modern Approach, Your Practice Rep — plus a hands-on lab, quiz, and project artifact.
Create a free account to unlock Phase 0 and Phase 1 of every course — no credit card.
Browse all courses · View pricing · DeVenture Academy