Prompt Engineering vs. Hallucinations
Crafting the Perfect Prompt for Each Model
Good morning!
Are you tired of vague and inconsistent AI answers?
This week's iteration is all about a video we recently shared for free, straight out of our AI for Business course, that breaks down the essentials of prompt engineering.
We show you exactly how to phrase requests so that ChatGPT (and newer reasoning models like o3) deliver sharper, fact-checked results, without wasting tokens on trial and error.
Spoiler: it's simpler than you think...
1️⃣ (sponsor) Generate a website in 30 seconds with this new AI website generator
Launch your dream business without spending $3000 on a website. You can edit all the pages, colors and sections!
2️⃣ ChatGPT Hacks: 7 Tricks to Instantly Boost Your AI Results
Let's go with this week's content!
Today, we cover the classics...
Zero-Shot Prompting – fire a plain request with no setup for quick wins.
Few-Shot / In-Context Examples – show 2–3 input-output pairs so the model mimics your pattern.
Chain-of-Thought Prompts – add “Let’s think step by step” to make the AI reason out loud for logic & math.
Role Prompting – assign a persona (teacher, comedian, analyst) to steer tone and depth.
System Prompts – hidden or top-level instructions that lock in tone, ethics, or response format.
Iterative Refinement – treat each run as feedback; tweak the prompt until the output matches your spec.
Hallucination Guardrails – ask for sources or cross-check facts to keep the model honest.
But not only the classics!
Spoiler — the best part: reasoning-first models like OpenAI o1, o3 can skip most “prompt hacks” if you front-load a rich, structured brief.
Here comes the importance of "context engineering", more than ever. It is much simpler than you think:
Want to mitigate hallucinations? It's two steps:
Get the best model possible for your budget or server.
Give it the best context possible. This means that it needs to have all the necessary information to answer and be as concise as possible.
This way, you are not asking it to "generate" an answer per se, but to "formulate" an answer from knowledge YOU control.
Then, you have one detailed prompt → one polished answer. Job done!
Watch the video for the full walkthrough and more details (or read the written version here):
3️⃣🚀 Ready to actually use AI at work—without needing to code?
If you’re a business professional who wants to lead with AI instead of watching from the sidelines, this course is for you. You’ll learn how to:
Use tools like ChatGPT, Claude, Gemini, and Perplexity effectively
Master prompting, reasoning, and no-code workflows
Automate research, writing, analysis, and team tasks
Create custom AI solutions that save hours every week
Drive smart, strategic AI adoption in your team or company
👉 Join the July cohort for our kick-off call of the “AI for Business Professionals” course here: https://academy.towardsai.net/courses/ai-business-professionals?ref=1f9b29
No technical background needed. Just real-world AI skills designed to boost your output, leadership, and career.