AI Assist Example—Fact Checking
ChatGPT might be useful for analyzing your content and letting you know if anything is wrong. However, use caution when doing this, because ChatGPT is not always correct. It is a good idea to cross-check its response with other sources.
Example You've written the following about Zilker Park in Austin, Texas:
You want to make sure it's correct. So with the text selected, you open AI Assist and ask ChatGPT to analyze it for correctness:
You receive this response:
On the surface, it might appear that ChatGPT has done its job. It correctly tells you that there are no volcanoes in Zilker Park, and therefore no recreational activity involving jumping into volcanoes exists. Good job, ChatGPT!
It also lets you know that your information is lacking some information. There might be other activities in the park that you are not listing. That might be helpful.
However, it does not catch the incorrect size of the park. Looking online at other resources, you discover that the park is only 351 acres, so you write the following:
At first, ChatGPT continues to give the wrong size:
You tell ChatGPT it is wrong, providing the correct number that you found from a number of other sources:
Finally, ChatGPT admits it was wrong:
In summary, you discover that ChatGPT might be useful for fact checking to a certain extent. However, you should not always trust its information. Instead, you should cross-check it with other sources.
