Here is a line directly from OpenAI’s documentation about ChatGPT:
“ChatGPT sometimes writes plausible sounding but incorrect or nonsensical answers…”
There are a lot of reasons for this and it has been getting better over time. The bottom line is - if the accuracy of the information matters, cross-reference with another source that you trust.
You can add instructions in your prompt to identify the sources that you want the AI tool to use. Unfortunately, this can also lead to incorrect information.
As the saying goes - “Trust but verify” (in this case I might skip the trust part and go directly to verify each time but that’s just me)
Comments