The following are a collection of publicly sourced articles related to the recent rise of artificial intelligence in everyday applications.
note: following each [hyperlinked] title/subtitle are abstracts generated by ChatGPT after receiving prompts from M. Neal. "Write an abstract for [article link]" is indicated with [URL] abstract. "Write an abstract for [article text]" is indicated with [text] abstract. These very similar prompt-responses are shared consecutively in the hope that reading the abstracts will showcase the variations in ChatGPT output (despite similar input).
How to Productively Address AI-Generated Text in Your Classroom
[URL] abstract: The article on "AI Generated Text" from Indiana University's Center for Innovative Teaching and Learning (CITL) discusses the increasing use of artificial intelligence (AI) technology in generating text and its potential impact on academic integrity. It highlights the importance of understanding the limitations and biases present in AI generated text and the responsibility of educators and students to critically evaluate and verify the authenticity of such text. The article provides practical tips and resources for identifying AI generated text and emphasizes the need for maintaining ethical standards in the use of AI in academic settings.
[text] abstract: “text too long”
How to Identify chatGPT stories?
[text] abstract: An editor shares insights on how to detect AI-generated text using GPT and offers alternatives such as observing telltale signs of machine-generated text, reading the story carefully, and comparing it to other sources. The editor notes that detectors are often wrong, and suggests using common sense and considering the value the text provides to the reader. The article concludes that the best way to identify texts produced by GPT is through layout and writing style, but the ultimate goal is to satisfy informational, entertainment, and emotional needs.
[URL] abstract: This article provides tips on how to identify AI-generated text produced by OpenAI's ChatGPT model. The article suggests to look for repetitive phrases, lack of detail, and absence of realistic human emotion as signs of machine-generated text. It also suggests using a plagiarism checker and comparing the story with other sources to check for inconsistencies and factual errors. The author acknowledges that detectors may not be entirely accurate and advises to use common sense and consider the value the text provides to the reader. The article concludes by emphasizing that what matters is how well the text satisfies our informational, entertainment, and emotional needs, rather than who or what wrote the text.Additional Resources for Identifying AI
The CheckGPT tool developed by Packback appears to be useful in guiding instructors on their responses to potential usage of AI-generated materials. More information about the tool can be found at https://labs.packback.com/originality.
last updated on 03/02/2023
