Generative Artificial Intelligence: Challenges and Possibilities of Generative AI

Generative AI is changing rapidly and it is important to keep in mind that content created by AI tools is the result of a predictive model; content is not necessarily accurate, unbiased, up to date, or ethical or legal to pass off as your own work.

Specific challenges of GenAI:

• Often "hallucinates" (makes up things) especially in terms of academic references or citations
• Has not been trained on the newest information, so has limited knowledge of current information and the current world 
• Does not have equal knowledge of all topics. For example, has more knowledge of STEM topics than humanities topics.
• Does not cite sources or otherwise indicate what in training data has led to the response
• As with search results, results must be critically evaluated for bias 
• Cannot make future predictions
• Lacks creativity or "out of the box thinking"

Always fact check content provided by an AI tool!

Keeping the challenges and limitations of GenAI in mind, here are things that generative AI tools can do:

• Write text based on specific prompts. Examples include: suggesting draft outlines for essays or research topics, drafting a formal email, or providing summaries of provided content.
• Create new visual media, such as an image
• Create new audio, such as a a music track
• Assist with writing code for a specific task

If you're going to use a GenAI tool or platform, be sure to use your critical thinking skills. This list of things to keep in mind, adapted from an NPR article from June 2023, is a good start:

Privacy: Be cautious about sharing any personal information with AI tools. These platforms may use your input for training AI models, and companies developing these models may have access to what you enter. Are you comfortable with your input data or prompts being shared?

Purpose: What are you using the software to create? Are you asking an image generator to copy the style of a living artist, for example? Or using it in a class without your teacher's knowledge? Consider the ethical implications of your use case and if you are unsure whether your use case is allowed in an academic context, refer to the UC San Diego standards for academic integrity.

Consent: If you're creating an image, who are you depicting? Could the person be harmed by the portrayal? What is your intent behind creating this image?

Disclosure: If you're sharing your AI-generated content on social media, have you made it clear this content is computer-generated? What would happen if they were shared further without that disclosure?

Fact check: Generative AI get things wrong and it is important to double-check any important information before you post or share it. Have you done this?