Skip to Main Content

Boatwright Memorial Library

Image that says "Defining Generative AI"

Where to start? Defining "Generative AI"

A key part of understanding generative AI, and specific tools like ChatGPT, is contextualizing it within the broader field of artificial intelligence. The following terms are ordered from conceptually broad to specific, and additional resources are included in each section: 

 

Image that says "Artificial Intelligence"

 

Image that says "Generative AI"

  • A specific branch of artificial intelligence focused on a machine model's ability to create, or "generate", new content. This content could be text-based, images, music, or even human voices. This is possible thanks to two key concepts:
    • Training data: the set of examples or input data used to train a generative AI model. It consists of a collection of representative samples the model learns from to generate new content or make predictions. For example, companies like OpenAI use web data collected by Common Crawl, a non-profit organization that has collected over 9.5 petabytes of web data since 2007.
    • Machine learning: an algorithmic process that identifies patterns in training data sets and makes predictions based on those patterns.
  • Additional Resources:

 

Image that says "Large-Language Models"

 

Image that says "ChatGPT" 

  • A specific chatbot LLM, created by the company OpenAI. ChatGPT can take a wide range of user text inputs and respond in a human-like manner. 
  • There have been multiple versions of ChatGPT made available to the public. The most well-known version, ChatGPT 3.5, was released in November 2022, though it has since been discontinued in favor of updated ChatGPT versions.
  • Additional Resources:

 

Source: AI Guide – The AI Pedagogy Project. (n.d.). Retrieved December 4, 2023, from https://aipedagogy.org/guide/

Image that reads "What is AI Literacy?"

Per a definition coined by Long & Magerko (2020), AI literacy refers to "a set of competencies that enables individuals to critically evaluate AI technologies; communicate and collaborate effectively with AI; and use AI as a tool online, at home, and in the workplace" (p. 2). 

While this is not necessarily an all-encompassing definition, and one that will continue to evolve alongside new generative AI tools, it speaks to key skills the UR Library will foster through this guide, and in our research assistance with students and faculty alike. In particular, we envision users will be able to:

  • Explain the basic functionality and applications of generative AI;
  • Utilize generative AI tools in an effective, responsible, and ethical manner;
  • Assess various generative AI tools, as well as their outputs, for accuracy and bias;
  • Analyze the various implications of increased generative AI usage, from privacy to environmental sustainability

For more information on AI literacy, check out the following pieces below:

Long, D., & Magerko, B. (2020). What is AI literacy? Competencies and design considerations. Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems, 1–16. https://doi.org/10.1145/3313831.3376727

Miao, F. & Shiohira, K. (2024). AI competency framework for students. UNESCO. https://doi.org/10.54675/JKJB9835

Ng, D. T. K., Leung, J. K. L., Chu, S. K. W., & Qiao, M. S. (2021). Conceptualizing AI literacy: An exploratory review. Computers and Education. Artificial Intelligence, 2, Article 100041. https://doi.org/10.1016/j.caeai.2021.100041

LinkedIn Learning Courses

What is generative AI? from Generative AI: Introduction to Large Language Models by Frederick Nwanganga

Find more information about generative AI through the following LinkedIn Learning courses (all UR students, staff, and faculty have access to LinkedIn Learning):

The LinkedIn Learning platform (formerly Lynda.com) is available to all UR students, staff, and faculty. 

Key Terminology

Here are terms commonly used in discussions of generative AI and their definitions:

Autonomous AI: Artificial intelligence systems capable of operating and performing tasks without constant human control. These systems can analyze data, learn from it, and execute actions based on their programming and algorithms.

Bias: The presence of systematic and undesired preferences or imbalances in the output generated by an AI model. Bias can emerge in various forms, such as in the content, language, or perspectives generated by the AI system.

Burstiness: The abrupt shifts in quality, coherence, or relevance often observed in AI-generated content, particularly in writing. It refers to the inconsistencies in style, tone, or factual accuracy that can occur within a short span. Identifying burstiness helps distinguish AI-generated content from human-created content.

Generative AI: AI systems that can generate new content, such as text, images, or music. It involves developing algorithms and models that can understand patterns in existing data and use that understanding to generate novel output.

Generative Model: An AI model designed to generate new data that resembles the patterns and characteristics of the training data it has been exposed to.

Hallucinations: Misinformation or made-up information based on a pattern that the AI model has learned as part of its training. For example, the model could create references that do not actually exist.

Heat Map: A visual representation that highlights important elements in the output generated by an AI model. It helps understand where the model focuses and assists in evaluating and improving the generated content.

Large-Language Models: Components of artificial intelligence developed based on the training of vast datasets of documents from various sources. The computer program analyzes data input and maps out words in the dataset. It next tries to predict which words are positioned before or after other words using predictive patterns of most likely combinations.

Output: The generated content produced by a generative AI system. It can be text, images, audio, music, video, or other data the model is designed to produce.

Perplexity: A measure used to assess the coherence and consistency of AI-generated text. Higher perplexity values suggest the content is more likely to be AI-generated due to unusual patterns or inconsistencies. Content identification systems use perplexity to identify AI-generated content.

Positional Encoding: A technique that assigns a number to each word during training that is used to show the position (or order) of words in a sequence.

Probabilistic: In generative AI, probabilistic means that the models incorporate probability, which is used to estimate the likelihood of different outcomes and generate outputs that align with the learned probabilities.

Prompt: The initial input text or instructions given to a model to generate new content based on that starting point. It provides context and guides the model's output. The prompt can be a few words or sentences that set the tone or specify the desired content.

Sentient: The capability to possess consciousness, self-awareness, and subjective experiences. Achieving true sentience in AI systems is a topic of scientific exploration and philosophical debate.

Tokens: Discrete units used to represent meaningful components of text, such as words or phrases. Breaking down text into these units allows AI models to process and analyze language at a granular level, enabling tasks like language generation. 

Training Data: Training data refers to the set of examples or input data used to train a generative AI model. It consists of a collection of representative samples the model learns from to generate new content or make predictions.

Transformer: A type of model (or robot) that can simultaneously work on several tasks and sequentially build output. The transformer gives AI models the ability to process and learn from data so they can interpret context and place words together to form a cohesive sentence structure

 

Adapted in part from: Walden University (n.d.) Key AI terms glossary. https://academics.waldenu.edu/artificial-intelligence/glossary

Additional Resources