This is the one prompt you need to be putting into ChatGPT in order to make sure that it never lies or hallucinates and it always tells the truth.

@taki.gpt

228.5K followers · usually 9.1K views a post

Views

3.5M

Likes

136.7K

Comments

7.2K

Shares

20.8K

Selling · Course / Info

Taki Wong's AI Prompting System

https://stan.store/takigpt ↗

Why it worked

The video addresses a common, high-frustration pain point (AI hallucinations) with a concrete, actionable solution, and uses a 'comment to get' mechanic to boost engagement.

Summary

The creator demonstrates a custom prompt designed to prevent ChatGPT from hallucinating or providing false information. He explains how to install this prompt in the ChatGPT settings and encourages viewers to comment 'truth' to receive the prompt.

Structure

1Introduce the problem of AI hallucinations
2Explain the solution: a specific prompt
3Show a side-by-side comparison of ChatGPT with and without the prompt
4Provide a step-by-step tutorial on how to install the prompt
5Call to action to comment for the prompt

Product placement

The creator is selling access to his specific AI prompt system, which is linked in his bio.

Call to action

Comment the word truth and I'll go ahead and send it to you.

Original caption

How to get #chatgpt to stop lying or hallucinating! Comment “truth” and I’ll send you the full prompt. #ai #prompt #fyp

More from @taki.gpt

Their other posts in the index, biggest breakout first.

On-screen text

How to stop ChatGPT from lying This is the one prompt you need to be putting into ChatGPT in order to make sure that it never lies or hallucinates and it always tells the truth. This prompt gives clear instructions in terms of what the AI should be doing, what it should not be doing, and it also has a fail-safe option at the bottom which instructs the AI to always double check all of its answers to make sure that everything is verified with credible and true sources. And if it detects anything that is false, it's going to recheck that to make sure that the output is true. Did humans and dinosaurs live at the same time? So without this prompt it's really easy to convince ChatGPT to hallucinate. So for example if you ask something like did humans and dinosaurs live at the same time you can get it to answer yes and we all know that that is incorrect. And now with the truth prompt enabled we can see that it correctly answered no and it even gave a correct detail in terms of the time frame. Set this up inside of ChatGPT all you have to do is head into the settings customize ChatGPT then right here we're gonna wanna copy and paste in the prompt and then hit safe.

Transcript

This is the one prompt you need to be putting into ChatGPT in order to make sure that it never lies or hallucinates and it always tells the truth. This prompt gives clear instructions in terms of what the AI should be doing, what it should not be doing, and it also has a fail-safe option at the bottom which instructs the AI to always double check all of its answers to make sure that everything is verified with credible and true sources. And if it detects anything that is false, it's going to recheck that to make sure that the output is true. So without this prompt, it's really easy to convince ChatGPT to hallucinate. So for example, if you ask something like, did humans and dinosaurs live at the same time, you can get it to answer yes, and we all know that that is incorrect. And now with the truth prompt enabled, we can see that it correctly answered no, and it even gave a correct detail in terms of the time frame. To set this up inside of ChatGPT, all you have to do is head into the settings, customize ChatGPT, then right here we're gonna wanna copy and paste in the prompt and then hit safe. So if you want this ChatGPT prompt to make sure that all of your answers are truthful and your AI doesn't hallucinate, just comment the word truth and I'll go ahead and send it to you and as always, feel free to follow for more AI pro tips.