Being Human: ChatGPT

Due to its popularity ChatGPT is often down or slow to generate responses during peak usage windows. This is important to know if  you are planning on using it in your class - you should have a plan b in case it isn't available at that time. It may be better as an out-of-class activity, as that will give students more time to generate the text. 

Current weaknesses:

• Accuracy, it is not design to produce factually accurate information
• Currency, as of January 2023 it has been trained on materials from 2021 and earlier
• Bias, it will replicated the biases in its training material
• Information behind paywalls, it may not have had access to certain information that isn't freely available

Resources

• "Assessment in the age of artificial intelligence" by Zachari Swiecki, Hassan Khosravi, Guanliang Chen, Roberto Martinez-Maldonado, Jason M. Lodge, Sandra Milligan, Neil Selwyn, Dragan Gašević in Computers and Education: Artificial Intelligence special issue: Empowering learners for the Age of AI.
• "AI Text Generators: Sources to Stimulate Discussion among Teachers" Compiled by Anna Mills for the Writing Across the Curriculum Clearinghouse as part of a larger resource collection: AI and Teaching Writing: Starting Points for Inquiry.

• "ChatGPT Guidance for the CUNY Classroom: Graduate Center experts weigh in on how professors can reap the benefits and avoid the pitfalls of ChatGPT." by Cathy N Davidson, Roderick Hurley, Luke Waltzer, Adashima Oyo, and Matthew K Gold

• "Adapting College Writing for the Age of Large Language Models such as ChatGPT: Some Next Steps for Educators" by Anna Mills and Lauren ME Goodlad

• "ChatGPT Advice Academics Can Use Now" by Susan D'Agostino for Inside Higher Ed.

Some conversations happening around ChatGPT that may be worth knowing as you consider using it as a tool in your own work/teaching:

• "Exclusive: OpenAI Used Kenyan Workers on Less Than $2 Per Hour to Make ChatGPT Less Toxic" by Billy Perrigo for Time. 
• "Training a single AI model can emit as much carbon as five cars in their lifetimes" by Karen Hao for MIT Technology Review
• "The lawsuit that could rewrite the rules of AI copyright" by James Vincent for The Verge.
• "How to Spot Robots in a World of A.I.-Generated Text" by Keith Collins for The New York Times
  A good explanation of how ChatGPT works
• "Why does chatGPT make up fake academic papers?" Twitter thread by David Smerdon
• "We Asked ChatGPT About Art Theory. It Led Us Down a Rabbit Hole So Perplexing We Had to Ask Hal Foster for a Reality Check" by Ben Davis for artnet
• "The Hidden Biases Behind ChatGPT" by Christian Ilube for People of Color in Tech