HD-OPM Archive
Topics › AI & digital tools

How to fact-check posts and comments with Google Gemini

Louis Nervegna · Mar 05, 2026 · ♥ 37 reactions · 11 comments · original Facebook post
AI & digital toolsGroup news & community
AI summary & key facts

Group-practice note encouraging members to fact-check posts, including the admins' own, using an AI tool and to flag inaccurate posts with evidence. Not health research.

One thing that makes this Facebook group different from others is that we try to fact check all the posts for scientific accuracy, including our own.

AI tools like Gemini now make this a lot easier.

We encourage everyone to do this.

If you feel that a post is not factually accurate, please report it, or point it out in the comments (with supporting evidence). If there was an error or oversight, we will try our best to get it fixed.

Fact checking posts and comments
Here’s how to fact-check posts and comments with Google’s Gemini AI:

This is completely free.

They don’t charge you anything to do this.

  1. Go to the Gemini website here:
gemini.google.comLink to gemini.google.comgemini.google.com
  1. Type in the following text, and paste what you want to fact check underneath:

“Please critique and fact check the following text for scientific accuracy. Do not include tables in your response:

<paste text here>”

———-

That’s it.

It’s important to start with a new clean session every time, so the AI is not biased by any past discussions that you had with it.

I do this on most posts and comments now, when I’m not completely certain about the factual accuracy— including my own.

I like Gemini because it provides clickable web links to the references that it uses to craft its response. That allows you to see exactly where it is getting its information from.

The part about not including tables makes the response easier to copy and paste into Facebok posts and comments. (The tables do not easily copy and paste without editing.)

Once you get in the habit of doing this, you’ll be amazed how much stuff you find on social media that looks legit at first, but turns out to be mostly nonsense.

Note: For this to work properly, the prompt has to specifically ask the Al to critique and fact check. That tells the Al that you want it to find things to disagree with. You want to avoid constructing the prompt in such a way that you encourage it to agree with you. Instead, you want to do the opposite; you want to encourage it to disagree with you.

#factcheckingHD #rulesHD #influencersHD #aiHD #evidenceHD

← newer: Master@Heart: higher wearable-measured training load linked older: Alzheimer's may start with inflammation in distant organs su →