This guide's intention is to help you engage with generative AI tools critically, responsibly, and ethically.

What chatbots are

Generative artificial intelligence (AI) is a rapidly evolving technology. Interactive AI tools (such as ChatGPT, Bard, Llama, Claude, etc.) represents a novel way to interact with information.

What chatbots do

A chatbot is a computer program that simulates human conversation by probabilistically generating word sequences based on its own analysis of a large amount of linguistic contents. In other words, chatbots respond to your prompts by guessing the answer. 

What chatbots are good at

Chatbots returns seemingly coherent responses. They can summarize large texts, debug programming codes, put together resumes, write essays, and answer multiple choice questions, among many other text-based capabilities. They can also alter, narrow, or expand on the response at the user’s request. Tasks chatbots are good at include summarizing, organizing, brainstorming, making unexpected connections, carrying on casual conversation, impersonating a particular speaker’s voice, and solving math problems when starting with symbolic expressions and equations. 

What chatbots are also good at, in a bad way

Since chatbots know only as much as what's in the training corpus, they are not designed to do things like telling the truth, making accurate and verifiable claims, making ethical judgements, and analyzing complex situations (such as math word problems), even if they insist with extraordinary eloquence and confidence. The derivative and probabilistic nature also makes it difficult for them to speak with a distinct voice or to be artistically creative at a high level.   

Chatbots for Research? 

Unreliable. You're better off using traditional tools for now.

Chatbots also are not equipped to write academic papers, because a large swath of academic literature behind the pay wall is not included in their training text. 

What do we do now?

1. Think critically
2. Use chatbots as a tool (to help but not to rely on)
3. Consider bias in AI
4. Confirm accuracy
5. Be transparent (with documentation)
6. Understand academic integrity
7. Ask about how to use them
8. Respect privacy and consent
9. Be ethical digital citizens
10. Spread the word about ethical AI practices

(Adapted from Carnegie Learning: https://www.carnegielearning.com/blog/ethical-ai-chatgpt-students/)

How to use chatbots critically, responsibly, and ethically as a university student.

https://chat.openai.com/share/5c4b0091-ed5a-44e6-8d43-ad8f925317be

Prof. Milch's assignment

https://chat.openai.com/share/fcc5855a-472d-45e7-b5a4-17fb77571327

You can follow my brief chatbot conversations here. I can tell you now that some of the references offered by the chatbot do not exist.