ChatGPT Enters the AI Debate
Launched in Nov. 2022, the artificial intelligence-enabled chatbot ChatGPT has caused a stir among supporters and detractors.
According to creator OpenAI: “We’ve trained a model called ChatGPT which interacts in a conversational way. The dialogue format makes it possible for ChatGPT to answer followup questions, admit its mistakes, challenge incorrect premises, and reject inappropriate requests. ChatGPT is a sibling model to InstructGPT, which is trained to follow an instruction in a prompt and provide a detailed response.”
A judge in Colombia, Juan Manuel Padilla, admitted in a radio interview that he used ChatGPT to draft a court ruling about whether a minor with autism is exempt from paying for therapies. He explained, the chatbot could “facilitate the drafting of texts” as a secretary might, but “not with the aim of replacing” judges. The judge noted, “I suspect that many of my colleagues are going to join in this and begin to construct their rulings ethically with the help of artificial intelligence.” The admission was met by some with horror. Juan David Gutierrez, a professor at Rosario University, calling for digital literacy training for judges, stated: “It is certainly not responsible or ethical to use ChatGPT as intended by the judge in the ruling in question.”
Tiera Fletcher, professional rocket scientist and co-founder of Rocket With The Fletchers, examined ChatGPT’s answer to “the rocket equation,” at the request of NPR. Tasked with describing the basics of what makes rockets fly, the chatbot “crashed and burned,” leaving out variables, failing to reproduce basic equations, and including errors.
The chatbot has raised debate among educators, with some banning and others embracing the tool. Educators are concerned about “AI-enabled plagiarism, academic dishonesty, and the propagation of misinformation.” More recently, perhaps in response to the education debate, OpenAI also launched an AI-detection widget that determines if text is “very unlikely to have been AI-generated,” “unlikely,” “unclear,” “possible,” or “likely.” However, OpenAI admits “Our classifier is not fully reliable. It should not be used as a primary decision-making tool.”
Discussion Questions
1. What are your thoughts on chatbots like ChatGPT? Are they useful? Problematic? Explain your answer(s).
2. Should ChatGPT and other AI text generators be banned in schools? Consider New York City Schools’ ban and a Wharton Business School professor embrace of the chat bot. Explain your answers.
3. How much transparency should journalists offer about using AI text generators? Consider CNET’s use of ChatGPT. Explain your answers.
Sources
Geoff Brumfiel, “We Asked the New AI to Do Some Simple Rocket Science. It Crashed and Burned,” npr.org, Feb. 2, 2023
CBS News, “Colombian Judge Uses ChatGPT in Ruling on Child’s Medical Rights Case,” cbsnews.com, Feb. 2, 2023
Lauren Leffer, “OpenAI’s New AI-Detector Isn’t Great at Detecting AI,” gizmodo.com, Feb. 1, 2023
