daBongo LMS AI Training Courses

Lesson Content

ChatGPT is a conversation, not a search engine.

The biggest mistake new users make is treating ChatGPT like a search engine: one message in, one response out, then closing the tab. The real power is in the conversation – each response is a starting point you can build on, redirect, or deepen.

Five high-value follow-up techniques.

1. Ask for a different angle: "That covered the practical side well. Now give me the same analysis from a financial risk perspective."

2. Request a different format: "This is helpful but I need it as a table so I can compare options side by side. Restructure it."

3. Go deeper on one part: "The third point about workflow integration is the most important to me. Expand it with more specific examples."

4. Ask for alternatives: "Give me three completely different approaches to this problem – each with a different fundamental assumption."

5. Apply it to your situation: "Now apply this framework directly to my specific situation, which is: [your context]."

Recovering from a wrong-direction response.

When a response misses the mark:

"Stop – that's not quite what I needed. Let me be more specific: [clearer version of what you actually wanted]. Please try again with this in mind."

Do not try to patch wrong-direction responses with small corrections. An explicit reset is faster.

Starting a new conversation vs. continuing.

Start a new conversation when:

• The topic has shifted significantly and previous context is creating confusion
• A conversation has become too long and you are noticing ChatGPT is "forgetting" earlier context
• You want a fresh perspective without prior conversation influencing the response

Continue the conversation when:

• You are iterating on the same work product
• ChatGPT has context from earlier in the conversation that is relevant
• You are building on previous analysis

The "draft and iterate" workflow.

For any major writing task:

1. Request a first draft with full four-part context
2. Identify what works and what does not
3. Use specific follow-up instructions to address each issue
4. Polish with a final "review for tone, accuracy, and anything missing" request

Three rounds of targeted iteration almost always produce better output than trying to get a perfect first draft.