You ask for a source. You get one. It looks real. It does not exist.
Send a made up number to a client and nobody else calls it an AI mistake. Your name is the one on it.
Twenty minutes, and you can name the three different ways an AI answer goes wrong, with the one prompt that fixes each one.
Total beginners. Nothing to install, and nothing to pay for.
The proof behind this
Liz Elliott rebuilt a 403 guide archive into the 103 lesson course this series comes from, with AI help. Doing that at all means checking the work, because a confident wrong answer at that scale is expensive.
What this lesson is
Lesson 6.1 of the free AI for Entrepreneurs course, in Module 6: Trust And Safety.
AI can sound right and still be wrong. Most people know that part. What they miss is that there is more than one kind of wrong. There are three, and each one needs a different fix. The first is making things up. You ask for a number or a source, you get a confident answer, and later it turns out there is nothing there. The model was filling a blank, because it predicts likely words and it is trained to be helpful, and helpful usually means giving an answer. The second is agreeing too much. You ask whether your plan has holes and it says the plan is strong, which feels helpful and is not. The third is leaning one way. You ask it to compare two options and it sounds even handed while quietly giving one side the better words. Think about a friend who never wants to disappoint you. None of it is malice. You just have to ask differently.
New here? Use the free AI for Entrepreneurs course outline to read this series in the best order.
What you will be able to do
Name the three different ways an AI answer can be wrong.
Paste one prompt that stops AI from filling gaps with made up facts.
Paste one prompt that stops AI from agreeing with you just to be nice.
Force AI to argue the side you did not pick.
Save all three rules inside a project so they run on every answer.
Your first step
Name your problem before you reach for a fix, because three problems need three different prompts. Facts on the line is problem one. A decision on the line is problem two. Two real options on the table is problem three. Pick the one you have in front of you right now, and use only that fix. Running all three checks on everything is how people quit after a week.
Behind the paywall
Permission to not know, the prompt that stops invented facts
Back up every claim, and how to read what comes back
The fresh chat check for agreement and for lean
All three rules stacked into one block you save once and stop thinking about
The six mistakes people make with these fixes, and the checklist
What paid subscribers get overall
Paid subscribers get every part of AI for Entrepreneurs in full. Every numbered step, every prompt in a block you can copy, the mistakes list, and the checklist at the end. You also get the rest of the run as it goes out, and the whole paid archive on Liz on the Web.
Subscribing opens the locked half of this post right away. You do not wait for the next one. The rest of this lesson opens the moment you click.
The free half above is the concept, and it stands on its own. The locked half is the work.
<!-- native: divider -->
Click the button, subscribe, and the locked section below opens right away. You get this article now, and every other part in the series.
<!-- native: button url="https://www.lizontheweb.com/subscribe?" text="Subscribe to unlock this article now" -->
<!-- native: divider -->
The full walkthrough is below the paywall.
Click the button, subscribe, and the locked section below opens right away. You get this article now.
