ChatGPT and other Large Language Models (LLMs), such as Google Bard, are powerful AI tools that can produce seemingly human-generated text using probability algorithms. If used correctly, these tools can assist you in various tasks. In fact, some of the text on this website was AI generated. However, to use ChatGPT (and other LLMs like it) in a well-informed and responsible manner, it is crucial to understand that what ChatGPT essentially does is to produce the most likely continuation of whatever text it has received up to that point, based on large amounts of text data from billions of pages of web sites and digitized books.
In other words, ChatGPT predicts what the next words or phrases in a sentence should be, to produce content that is meaningful to us humans.

​

Thus, while ChatGPT can be a valuable tool, there are several caveats and considerations to keep in mind:

• ChatGPT is not reliable. ChatGPT generates responses based on patterns learned from large datasets of text from the internet. This means that ChatGPT is only as good as the data it is trained on. If the training data is not diverse or representative enough, its outputs may be incorrect, incomplete and/or inappropriate. It is vital to realize that ChatGPT lacks the ability to verify the information it presents or evaluate its reliability. Therefore, if the facts matter,  you should always verify the information ChatGPT provided, as shown in the following flowchart.