Do you agree with this perspective on ChatGPT? Citation in GoogleDoc link. https://docs.google.com/document/u/0/d/1rBDP4WsMqv8bIw6tK6-SZ0sDh0_OcperWMcabn5II9Y/mobilebasic #education #edtech @edutooter @edutooters
@mguhlin relatively Torrey Trust is fantastic and I wish she would come to Mastodon
@mguhlin @edutooter @edutooters Totally agree with it. Thanks for this great overview! Maybe the problem is specific to some subjects. In science, I'd rarely ask students to do “googleable” tasks. It's more about analyzing data, elaborating concepts and so on. But the overview offers some possibilities to handle even simple repetitive tasks.
@herr_sander @edutooter @edutooters To be clear, I didn’t come up with it. The author is on the document. That aside, it was interesting and worth sharing. I haven’t formed an opinion yet. 🙂
@herr_sander @mguhlin @edutooter @edutooters at my Uni, we have come up with a similar set of flags for our students, emphasising both the LLM’s lack of attribution of the information it uses and of some of the pitfalls of trusting it with your information OR assuming it has the correct information. We have done a bit of playing, and it looks like (until it learns better!) it often makes up journal articles to cite while using actual journal names, for example.
@jnyrose @herr_sander @edutooter @edutooters That’s hilarious! Thanks for sharing your experiences.
@mguhlin @edutooter @edutooters 100% with this. I especially appreciate the point about assigning "Googleable" work. I think one of the things that Google searches to ChatGPT ask of us as educators is what student worthy tasks are. I have not worried about "cheating" in years because virtually all of the tasks we ask students to do involve analysis, revision, and solving complex challenges without simple or straight forward answers.