In winter 2023 we talked with Andy Jones and Margaret Merrill of UC Davis about ChatGPT, a new artificial-intelligence app that was setting off alarms for its advanced ability to "write" reports and articles. On today’s Davisville they report that among the faculty they work with, the sense of panic present then has now eased “quite a bit.” People know more about the limits of chatbots, and are asking more about how and where to use the tools in teaching, instead of just fearing them as a plagiarism machine.

We talk about handling chatbot hallucinations, resisting the biases that chatbots suck into their text databases, and hear a few examples of how UC Davis instructors are using the tools in their classrooms.

Too many AI tools are arriving too quickly to keep up with them all. You have to keep up, but not with every twist and turn. “Rather than always trying to figure out what is in store for us from the next AI engine,” Jones advises, look for ways to “connect meaningfully with one another.” The human connection is essential.

(This image wasn’t created by a chatbot, but the caption error illustrates the hallucinations the bots can create.)