Artificial Intelligence

• ChatGPT and others are large language models (LLMs) that are "trained" on large amounts of data pulled from all over the world from huge databanks.
• They generate statistically probable text from that data, creating new work similar to, but not identical to the original data.
• LLM's do not search for information and retrieve the original input for you to consider.  They create new text derived from the training data.
• They don't provide citations. 
• There is no way to reproduce the results.  You will get something different every time, even with the same prompt.  Similar, maybe, but not the same.
• There are major issues with copyright and user privacy.    
• Be very skeptical of the results.  Do not trust any information you cannot verify yourself using other sources.  ChatGPT and other large language models have well-known examples of "hallucination" (fake information) or "ghost citations" (fake sources).
• Be scientific. Can you replicate the information provided?  The same prompt at a different time may result in a difference response from the GenAI you are using.  Changes in wording may yield different responses.  Try a different large language model.
• Be aware of bias.  To "train" the system, generative AI ingests data from across the Internet.  Therefore, AI can replicate the biases, stereotypes, and hate speech found on the web.

Machine Learning

Machine learning is a subfield of artificial intelligence (AI) that focuses on developing and studying statistical algorithms that can learn from data and generalize unseen data.  It allows machines to perform tasks without explicit instructions.

Neural Network

A neural network is a subset of machine learning and consists of layers of nodes, or artificial neurons - an input layer, one or more hidden layers, and an output layer.  Each node connects to others and has its own associated weight and threshold...One of the best-known examples of a neural network is Google's search algorithm. 

Large Language Model

A large language model (LLM) is a type of machine learning model that can comprehend and generate human language text.  These models work by analyzing massive data sets of language, allowing them to achieve general-purpose language generation and perform various natural language procession tasks.  LLMs are revolutionizing chatbots, virtual assistants, content generation, and language translation.

Generative AI

Generative AI refers to deep-learning models that can generate high-quality text, images, and other content based on the data they were trained on.  Generative AI can take raw data (like the complete works of Shakespeare) and "learn" to generate statistically probable outputs when prompted.  Generative AI can then draw from the training data to create a new work that is similar, but not identical to the original data.

Hallucination

An AI hallucination is where a large language model perceives patterns or objects that are nonexistent or imperceptible to human observers, creating outputs that are nonsensical and/or inaccurate.

Definitions above were generated by Microsoft Copilot and fact-checked by a UHV librarian.

For more information about Artificial Intelligence terminology...

10 AI terms everyone should know.  Susanna Ray/Microsoft.

10 (more) AI terms everyone should know Susanna Ray/Microsoft

Wikipedia Glossary of Artificial Intelligence

Artificial Intelligence Definitions. Human-Centered Artificial Intelligence.  Stanford University.