I’m giving a free Zoom webinar tomorrow (2 November) at 6pm UK time on how to evaluate Generative AI tools for scholarly texts. The talk is part of the MAL education series. I don’t claim to have all (or even most of) the answers, and certainly not a full understanding, but clearly it’s worth having a discussion about how to use Generative AI without too much hyperbole.
A couple of years ago I gave a talk at UKSG on “An AI Toolkit for Libraries”, which became an article, and this talk is an attempt to see if there could be a set of relevant criteria for judging ChatGPT and the other generative tools when used some defined contexts – for scholarly texts. My approach is that the best way to make use of the new tools is to understand something about their background, and so have a better idea about what they might do well, and what not so well. It might just be that you have to ask the right questions! They might be black boxes, but we have learned from our experience as users how to make effective use of AI tools before, and so there should be some ground rules should apply in the same way. There will be plenty of time for comments and discussion, and I look forward to hearing what others think!