ChatGPT (Chat Generative Pre-trained Transformer) has caused a stir across many professions and has been a hot topic of debate across universities and publishing houses. The tool has exceeded expectations and even passed academic exams raising several ethical considerations about ownership, authorship, and plagiarism. We discuss these considerations and how the model was trained to be effective and politically correct, which also raised further critiques on potential biases and the method of training.
Debate participants:
- Stephanie Forkel, Ass. Professor at the Donders Institute, Netherlands
- Lilit Dulyan, PhD student at the Donders Institute in the Netherlands
- Chris Foulon, Postdoc at UCL in the UK
- Michel Thiebaut de Schotten, Research director at CNRS in France
- Barbara Molz, Postdoc at the Max Planck Institute, Netherlands
- Ann-Katrin Ohlerth, Postdoc at the Max Planck Institute, Netherlands
- Sabrina Beber, PhD student at the Donders Institute in the Netherlands
- Marius Braunsdorf, Postdoc at the Donders Institute in the Netherlands
- Barbara Eckstein, Donders Institute in the Netherlands
- Joanne Kenney, Senior Clinical Scientist, UK
- Eva Guzman Chacon, Master's student at Radboud University, Netherlands
- Anna Matsulevits, PhD student at the University of Bordeaux, France
- Ambra Ferrari, Postdoc at the Max Planck Institute, Netherlands
- Giacomo Bignardi, Postdoc at the Max Planck Institute, Netherlands
- Danny Thewissen, School teacher in the Netherlands
- Valentina Pacella, Ass. Professor at the University of Padua, Italy
#teaching #research #innovation #ai #techinnovation #chatgpt #ethics #trainingdata #rewardlearning #DAN #publishing #universities #cnsdebate #cnsnightcap
---- Chapters -----
0:00 Introduction
2:09 Would you pay for ChatGPT?
8:30 What's under the hood: a reward learning model
9:35 Ethics and training datasets
9:52 Political Correctness: Trump vs Biden
11:02 DAN roleplay
12:04 What jobs might Chat replace
15:08 Editorial debate about plagiarism
16:48 Grammarly vs chat
21:34 Journal cover letter
25:55 Interim summary
32:14 Teaching innovation using AI
51:55 GPT competitors
