Artificial Intelligence Fundamentals and Applications

Comprehensive vocabulary flashcards covering the fundamentals, history, models, computer vision, natural language processing, applications, and ethical considerations of artificial intelligence across Chapters 1 through 6.

Generative AI

A type of AI that discovers the rules and patterns of its input training data to generate new data—such as text compositions, realistic images, high-quality video, original audio, or programming code—that resembles and builds on the original training content.

AI Chatbots

Computer programs using generative AI to interact with customers in a way that mimics human interaction, offering immediate and personalized answers to customer questions.

AI Algorithm

An advanced set of instructions that can improve a machine's performance over time so that the machine gets smarter from what it learns.

Collaborative Robots (Cobots)

Robots in manufacturing designed to assist human workers or other robots in tasks such as assembly, material handling, and quality inspection.

Autonomous Mobile Robot (AMR)

A type of robot that navigates a manufacturing plant using advanced sensors and navigation systems to autonomously perform tasks such as quality control and inventory.

Language Model

An AI model that learns from collected text data to produce original text by predicting the next word in a response, capable of recognizing voice, optical characters, and handwriting.

Generative AI Prompt

Natural human language instructional text that serves as the starting point for performing a specific task, such as drafting a business email or writing a resume.

Reinforcement Learning from Human Feedback (RLHF)

A training method that combines reinforcement learning techniques—such as rewards and comparisons—with human guidance and feedback to help an AI avoid common pitfalls and generate appropriate responses.

Hallucinations

Misleading or incorrect responses generated by ChatGPT and other AI tools that must be critically evaluated and corroborated with other sources.

Turing Test

A benchmark proposed by Alan Turing in the 1950s proposing that a machine can be considered intelligent if it can hold a conversation indistinguishable from one with a human.

Machine Learning (ML)

A concept introduced in the 1960s where computers learn to recognize patterns from many examples, make decisions, and learn from those decisions without being directly programmed with specific rules for every decision.

Perceptron

A basic model for an artificial neuron inspired by the structure of the human brain that receives data and filters it into two categories.

AI Winter

A period of reduced funding, interest, and progress in the field of AI—occurring in the 1970s and late 1980s—during which researchers reevaluated their work due to technology being insufficient for true AI.

Computer Vision (CV)

A field of artificial intelligence that allows computers to perceive, analyze, and interpret visual data from the world visually as a human does.

Keyword Spotting

An early NLP approach that detects specific keywords or phrases in audio recordings and typed text to locate matching predetermined coded responses.

Deep Learning

A type of machine learning that teaches computers to handle data in a way that resembles human brain functions, typically structured with input, hidden, and output layers.

Copilot for Microsoft 365

A deep learning-powered virtual assistant that uses prompt inputs to assist users with brainstorming, researching topics, developing drafts, and enhancing documents and presentations.

Artificial General Intelligence (AGI)

A type of AI that can understand, learn, and apply knowledge indistinguishably from human intelligence, distinguished by its ability to solve multiple problems in new situations rather than a single task.

Natural Language Processing (NLP)

A branch of AI that empowers computers to understand, process, interpret, and generate human speech or text responses.

Text Preprocessing

The phase in which an AI cleans and filters raw prompt data by lowercasing text, tokenizing, removing irrelevant words, stemming, tagging parts of speech, identifying named entities, and analyzing context.

Alt Text

Automatically generated descriptive text for images created using NLP to make visual content accessible to individuals with disabilities.

Google Gemini

A multimodal AI system that accepts voice and text inputs to process and generate text, code, audio, and images by analyzing scenes and recognizing objects similarly to human reasoning.

Big Data

Massive volumes of data stored in datasets within large data centers, an era triggered by the widespread adoption of the Internet in the 1990s and used for analytics, research, and business decision-making.

Large Language Models (LLMs)

Advanced AI models trained on vast datasets to comprehend and generate human-like text, originating from early programs like Eliza and advancing to modern tools like ChatGPT.

Claude AI

An LLM developed by Anthropic known for its safety and efficiency, integrated into Amazon's Alexa tool to enhance conversational skills and task performance.

Apple Intelligence Platform

A personal intelligence system integrated into iPhone, iPad, and Mac devices designed to enhance writing efficiency and creativity.

Scene Understanding

A computer vision capability that analyzes and interprets relationships between objects, spatial relationships, and activities within an image to comprehend the overall context.

Diffusion Model

A type of generative AI model that learns patterns and structures within large datasets of existing artwork to create new artworks, voice recordings, and movies.

Style Transfer

The process of transferring the distinctive artistic or visual style of one image into another AI-generated image.

Core Components of a Text-to-Image Prompt

The six essential elements of an effective image prompt: an action word, subject definition, style, scene, color palette, and technical specifications.

Black Boxes

A term describing AI systems whose internal decision-making processes lack transparency, making it difficult to understand or evaluate how they arrive at decisions.

Intellectual Property

Creations of the mind that receive legal protection against unauthorized use by others.

Copyright Infringement

The use of copyrighted work without obtaining the owner's permission.

Fake News

False or misleading information presented in the same format as legitimate news, produced and distributed by a fake news purveyor.

Semantic Scholar and Powerdrill

Specialized business and scientific research AI tools used as alternatives to general LLMs to discover relevant papers, extract key findings, and identify emerging market trends.