AI Key Vocabulary

HOW AI WORKS

A Simplified model

TRAINING DATA

DEEP LEARNING

NEURAL NETWORK

GENERATIVE AI

Training Data

All AI systems begin with very large sets of training data. This data may consist of text, images, audio, video or some other kind of media.

• The larger the data set, the more robust the generative AI will be.

• Bias in the training data can cause bias in the generative AI tool.

Deep Learning

AI systems use machine learning to find patterns within the training data and make inferences, not unlike how humans learn. Most systems use a type of machine learning called deep learning that organizes inferences and patterns into a neural network.

• A Large Language Model (LLM) is a type of deep learning process.
• A GPT (Generative Pre-Trained Transformer) is a type of LLM developed by OpenAI that can be customized for a variety of contexts.

Neural Network

A neural network is a structure that is created as a result of deep learning. It connects concepts and data points together similar to how the neurons and synapses work within our human brains.

Generative AI

A user gives prompts to a Generative AI tool that uses the neural network to create an original artifact. The Generative AI tool ingests the prompts and user feedback into its training data and uses it to improve the neural network as part of the deep learning process..