Using AI to Persuade Us

Thanks to the Genealogy and Artificial Intelligence Facebook group—and encouragement from Randy Seaver—I’ve become a frequent user of AI tools. I primarily rely on them for transcription, editing blog posts, and turning genealogy reports into narratives. Because of these experiences, I believe that learning to use AI responsibly should be a basic educational goal for all of us.

Recently, I’ve noticed something curious on my own Facebook feed. Posts keep appearing that I never liked, joined, or followed. They’re not advertisements, either. Instead, they seem to be content Facebook’s algorithm has decided I need to see. Many are about historical topics or the Kansas City Chiefs. At first, I found them interesting and even shared a few. But the more I read—especially the Chiefs-related ones—the more I started questioning their accuracy.

For example, one post claimed that Patrick Mahomes had traveled to Minnesota to place flowers at the memorial for children killed in the Annunciation Catholic School shooting and had even paid for their funerals. Knowing Mahomes is generous and active in charitable work, the claim didn’t seem far-fetched. But I also wondered how he could be in Minnesota while still attending practices. That curiosity led me to do some fact-checking.

Remember when fact-checking was common practice? People would ask, “Did you check Snopes?” I didn’t check Snopes.com this time, but I did ask ChatGPT. The result: nothing could be found to verify the story. That was enough for me to start clicking the X to hide these kinds of posts.

Another example involved a supposed quote from Travis Kelce.

While he may or may not have said it, once again ChatGPT could not confirm any source for the statement.

I’m increasingly convinced that many of these posts are being generated by AI and shared with the intent to persuade—or mislead—us. That’s why it’s important to dust off some old skills and apply them in this new environment:

• Fact Check: Use Snopes.com, PolitiFact, or similar tools to verify questionable claims.
• Quality Check the Source: As genealogists, we know that a marriage license is stronger evidence than a family tree entry. The same logic applies to news: a report with clear sourcing carries more weight than “someone said ….”

AI tools can be powerful for research and writing. But they can also be used to distort and manipulate. It’s up to us to recognize the difference—and to keep questioning what shows up in our feeds.