The use of artificial intelligence in writing is dramatically changing higher education. As someone who studies the impact of computing on literature, I am particularly aware that the writing process — as we’ve known it since at least the invention of the printing press — is changing forever. Many of my colleagues are concerned about the possibility for cheating among students, but it’s worth recognizing that these concerns are part of larger historical transformations that won’t go away any time soon.
ChatGPT (GPT refers to “Generative Pre-Trained Transformer”) is one of what are known broadly as LLMs, or large-language models. Also called generative AI, these are probabilistic models of natural language processing that generate text based on the statistical probability that one word will follow another. Another, less-well-known example of an LLM is Google Bard. It is worth noting that LLMs have many benefits. For instance, many students with disabilities, who speak languages other than English, or who are first-generation and unfamiliar with the conventions of academic English can find new forms of support with LLMs. Further, there are many ways ChatGPT could be used to supplement creative approaches for writing; for instance, writing in the style of a particular author.
I do not prohibit the use of ChatGPT in your writing. However, I will look for evidence of original thought in your work and grade against you if you use AI to replace your all of your thinking and interpretive work during the semester. I suggest that you document all use of ChatGPT when you use it. One way to do this would be to add a link to the interaction transcript that ChatGPT generates when it is used. Consider that the most important part of reading and writing is processing and developing an intellectual response to the works we read, thereby developing what pedagogue and activist Paulo Freire calls “critical consciousness.” That’s very difficult to do if an AI is doing all of the work for you. At the very least, you deprive yourself of the opportunity to develop your critical awareness of the world; yet, you may also be manipulated by the proven bias inherent in algorithms. See for instance, the work of Cathy O’Neil, Ruha Benjamin, and Shea Swauger. I would particularly recommend the 2020 film Coded Bias, for those of you who want to know more about algorithmic bias.
Consider the following facts about ChatGPT:
- To create an account for OpenAI, you must provide personally identifiable information like an email/phone number/or Google Account.
- OpenAI’s terms are quite clear that they collect and use your data as well as sell and/or share your data with third parties.
- When you use ChatGPT, you are providing free labor to OpenAI’s product development. They mention this on their FAQ page.
- It is quite possible that AI will dramatically change the job landscape in the near future, including jobs you may want to have.
- OpenAI exploits global labor, often traumatizing low-paid workers who are forced to review disturbing passages of hate speech and violence as examples to help the AI prevent the use of such speech on the app. Much of this work is done by refugees or people living in Kenya or the Phillipeans who will accept the $1-4/day salary offered by OpenAI.
Here are some good sources for those of you wishing to read more about the issue. Many of the recommendations mentioned in this policy statement come from the MLA/CCCC Joint Task Force on Writing and AI Working Paper along with Maha Bali’s research on Critical AI Literacy and Leon Furze’s work on AI Ethics.
Roger Whitson 