Using AI to Evaluate a Competency-Based Online Writing Course in Nursing from Online Learning
I’ve been bumping into this expression more and more recently about GenAI - it runs along the lines of ‘why should I bother to read something that someone couldn’t be bothered to write?’ At first glance it rings true and it certainly seems applicable in this instance - why should a student bother to create work for a teacher who won’t even read it? I don’t know that this article answers that question but it does offer a fairly detailed explanation of work undertaken to use GenAI to provide nursing students with automated feedback in a course about scholarly writing. (Why do most nurses need to learn scholarly writing vs healthcare writing - that is another question). What matters is that PandorAIs box is open on a lot of these issues and we at the very least need to be aware of what the outcomes are. (Overall, slight improvement, for the record)
Ending Human-Dependent Peer Review from The Scholarly Kitchen
Peer review in academic publishing is fraught, to say the least. Billion dollar publishers exploiting scholarly obligation, egomaniacs taking out their frustrations with the world on anonymous authors, a desperate shortage of reviewers causing bottlenecks in academic publishing, and so on and so on. This thoughtful piece from Haseeb Irfanullah steps through the many challenges of peer review and asks some decent questions about why we are not already using GenAI. After all, ChattyG probably isn’t going to ask you to cite their unrelated work in your paper.
It’s not like a calculator, so what is the relationship between learners and generative artificial intelligence? from Learning: Research and Practice
Another thoughtful piece on GenAI, this time from Jason Lodge, Suijing Yang (both UQ), Leon Furze and Phillip Dawson (both Deakin). As we try to come to terms with how this is changing our relationship with technology from controller to some kind of collaborator, ideas around the location of our ‘mind’ vs our physical brain and what we now need to do with either are surfaced as the authors consider cognitive offloading, the ‘extended mind’ hypothesis and new ways of learning entirely.
Olilo Multi-LLM Querying System from Olilo
Wrapping up this little GenAI deep dive, this handy little tool allows you to see the output of four LLMs to a prompt in the one window. Currently supporting ChatGPT4, PALM2 (Google), Claude (Anthropic) and Jurassic 2 Ultra (AI21 Labs), with LLama2 (Meta) and StableLM coming soon. I asked a simple question about flipping rows and columns in Excel and was a little underwhelmed by the response - particularly seeing that ‘I’ had run out of ChatGPT credits - but this looks like a site to come back to.