View Transcript
Episode Description
In this episode of Under Review, I dive into a recent and thorny side of the academic ecosystem, that being AI hallucinations. While Large Language Models are revolutionising research, they are also ‘bullshitting’ their way into journals, fabricating references, and spreading like weeds through the publishing landscape.To get to the heart of the problem, I decided to put one of the perpetrators under the spotlight. This experimental episode features an AI-generated interview response to my questions with ChatGPT, voiced by a custom AI-voice tool to reflect the tech bro culture behind the code. Whilst those listeners who are more tech-savvy will be aware of the issues, they may be less aware of how far we are from fixing this issue, especially when some academics and students are falling over backwards to publish more and more AI-written papers and images. AI-generated voices are still far from perfect and this might not be everyone’s cup of tea, but an experimental approach to putting one of leading generative AI tools under review makes for an interesting listen. This episode uses AI-generated responses to surface systemic issues within academia, reflecting broad debates and critiques rather than personal lived experience.The transcript from the interview can be viewed via this link.https://docs.google.com/document/d/e/2PACX-1vRsHlBQZfle0e_-7B6XkPwjLFuvpQdr3m6ryOo-es04jYTFEz5i3QMu20LzprvX8YC7zXewfyj_z2p_/pub
Hosted and Produced by Andy Tattersall
https://www.andytattersall.com/
I’m Andy Tattersall and I’m a research communications consultant with 25 years previously working at a Russell Group university and over the course of this podcast I will speak to some of the sector’s boldest thinkers to bring us their most radical, left-field and blue sky ideas for reform.
Music courtesy PineAppleMusic