In addition to evaluating the tools you use with the ROBOT tool, it's crucial to assess the material that the AI tools produce in response to your prompts. These tools are probabilistic and can often produce hallucinations that sound confident and informative, while being fabrications that can wrong. In libraries, we regularly see these tools invent citations for publications that look real, but don't exist.
Here are some questions to consider when evaluating output from generative AI tools:
You might notice that these questions are difficult (or sometimes even impossible) to answer when using generative AI tools. You will have to decide how this affects if and how you use the information you get from these tools.