AI, ChatGPT, and more generative tools: Pros and Cons of AI and ChatGPT

General Concerns

• As ChatGPT Enters the Classroom, Teachers Weigh Pros and Cons

• ChatGPT and Generative AI: Key Legal Issues

• Ethical and Privacy Concerns | chatGPT and AI

• ChatGPT Confirms Data Breach, Raising Security Concerns

• AI Platforms like ChatGPT Are Easy to Use but Also Potentially Dangerous (Scientific American)

• Ethical and social risks of harm from Language Models

• On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?

• Pause Giant AI Experiments: An Open Letter

• What Exactly Are the Dangers Posed by AI? (New York Times)

Opportunities and Challenges

• Is ChatGPT a threat to education?  

• Educator Considerations for ChatGPT (OpenAI) 

• Why ChatGPT is an opportunity for schools (International Baccalaureate)

• ChatGPT and Beyond: How to Handle AI in Schools (Common Sense Education

• The Ethics of College Students Using ChatGPT

• Embracing the Use of ChatGPT in Education  

• Some educators embrace ChatGPT as a new teaching tool   

• How to use ChatGPT as a learning tool

• Navigating the Risks and Rewards of ChatGPT

The dangers and limitations of generative AI

There are several limitations and concerns with the usage of generative AI.

• Copyright Infringement - Because AI bots get their data from unnamed sources, content creators work may be unwittingly mimicked or plagiarized. 

• Bias - We cannot know where generative AI tools get their data from or how they are using the data. Biased information is rampant on internet, so text created with AI tools is prone to the repetition of these biases. 

"Humans are Biased: Generative AI is Even Worse." YouTube, uploaded by Rajistics, 26 Jul 2023.

• Hallucinations - AI tools can make up incorrect outputs. A hallucination can occur when an AI bot creates a source that does not exist. All factual information will need to be researched from reliable sources thoroughly for accuracy. 

"Ai hallucinations explained." YouTube, uploaded by Moveworks, 18 May 2023.

• Inability to replicate emotions and voice and specificity - AI-generated text (currently) cannot replicate a consistent writer's voice on the page, nor does it understand human emotions. For the purposes of this class, you need to work with specificity, and sensory details, and develop a writer's voice. 

• Disinformation - Generative AI has the potential to amplify disinformation (intentionally false information) online by producing realistic fake content (such as deep-fake videos) that can undermine trust and spread falsehoods.

• Citations - AI-generated text cannot relied upon for the creation of accurate citations. As AI bots hallucinate sources and make up fictitious information, citations produced by AI-generated text must always be fact-checked. 
• Limit access to recent data - the current free versions of ChatGPT, Bard, and other AI bots haven't been trained on any data after September 2021, so doesn't include information on events or happening within this time-frame.

-

Some of the above content was derived from Nikia Chaney, Cabrillo College, via Canvas Commons

Best Practices with AI-Generated Content

Generative AI is an incredible and powerful tool, but it is essential that it is used in an ethical manner. Some ways to use it with your school work are as a tool for helping with brainstorming, outlining, and structure.

Since rules often vary from class to class you'll always want to check with your instructor (or check the syllabus) to see if there is a specific policy about using AI tools in your class.